Journal Issue

Scope and Limits of International Economic Cooperation and Policy Coordination

International Monetary Fund. Research Dept.
Published Date:
January 1988
  • ShareShare
Show Summary Details

Considerable attention has recently been directed toward the potential benefits of increased policy coordination among the larger industrial countries to improve the functioning of the present exchange rate system and to sustain world economic growth and price stability. Recent attempts to lay the groundwork for coordination include the 1985 reports of the Deputies of the Group of Ten and of the Group of Twenty-Four (see IMF Survey, Vol. 14, Supplements of July 1985 and September 1985, respectively), the Tokyo and Venice Economic Summits in 1986 and 1987, and recent meetings of the Interim Committee of the Board of Governors on the International Monetary System (henceforth, “Interim Committee”) held at the International Monetary Fund.

The Fund has contributed to the process through its surveillance over the policies of member countries. The analysis of the Fund’s biannual World Economic Outlook exercise has for some years emphasized economic developments in the major industrial countries whose policies and performance exert a powerful influence on the operation of the international monetary system and has included a discussion of the interactions of national policies of these countries and their repercussions on exchange rates and payments imbalances.

The purpose of this paper is to consider international economic cooperation, defined to include sharing of information and international surveillance, as well as the more specific issue of policy coordination among governments. The analysis draws on existing literature, both academic and more policy-oriented studies, and also reviews the actual operation of the present international monetary system from this perspective. In doing so, it attempts to throw light on several important questions:

  • What is meant by the concept of “policy coordination,” and how does it relate to other terms such as “cooperation,” “convergence,” and “harmonization”?

  • What is the relationship between the rules governing the international monetary system and the need for policy coordination, and the ability to achieve it? In particular, how does policy interaction differ under fixed and floating exchange rates?

  • What are the possible welfare gains from policy coordination under the current system of flexible exchange rates?

  • What policy lessons might be drawn from the experience with coordination under the present international monetary system?

  • How can the effectiveness of international cooperation be improved?

The central point of the theoretical literature on policy coordination is that there may be “externalities” relative to a country’s macroeconomic policies that imply that policies chosen optimally from an individual country’s viewpoint are less than optimal in a global context. An example may be the use of the exchange rate as a competitive instrument, either to achieve output growth by excessive depreciation (as in the 1930s) or to decrease inflation quickly through exchange rate appreciation. If all countries simultaneously attempt to depreciate their exchange rates or to increase their current account balance, these separate strategies will be self-defeating and will lead only to overinflationary or overdeflationary world policies.

Furthermore, there is a considerable lack of understanding of the effects of policies, both domestically and internationally, on the part of policy-makers, and the authorities in a given country usually must make decisions under conditions of uncertainty about the policy intentions of foreign countries. Cooperation in the form of exchange of information is necessary to permit countries to choose policies that are in their self-interest. For instance, a country may embark on a fiscal policy that will ultimately prove unsustainable but whose consequences, both at home and abroad, are not initially appreciated. There is therefore a role for identifying problems of unsustainable domestic policies and incompatibility of national objectives and policies in an international forum.

Concepts of Cooperation and Coordination

It is important first to be clear about terminology. Several closely related concepts—cooperation, coordination, convergence, and harmonization—have frequently been used interchangeably in the literature but have sometimes been defined in different ways by various authors. Interdependencies between countries arise for a number of reasons, including trade and financial links that transmit the effects of a policy-induced disturbance from abroad or of an external shock on key macroeconomic variables in the domestic economy. Similarly, feedback effects from other countries, including policy reactions to an initial disturbance, may modify the response of these variables to a policy action taken domestically—and potentially frustrate the achievement of domestic policy goals. Narrowly defined, and as used in the game-theoretic literature, policy coordination refers to decision making that maximizes joint welfare and thereby enables these international interdependencies to be positively exploited.1 This is intended to contrast with an “uncoordinated” regime, whereby each country is assumed to maximize its own welfare independently. A need for coordination in the narrow sense exists if the independent pursuit of national policies results in global outcomes that from the world point of view are nonoptimal. Among policymakers, the term policy coordination has long been used in a broader sense, to refer to agreements between countries to adjust their policies in the light of shared objectives or to implement policies jointly.2 The following analysis provides a general discussion of key macroeconomic policy issues facing the major industrial countries; therefore it frequently uses the broader definition of coordination employed by the policymakers. The narrow definition, however, is used where applicable, particularly in discussions of the academic literature, as in consideration of the welfare implications of game theory.

The term cooperation is here used in a very broad sense to encompass all forms of interchange between countries that relate to economic developments and policy intentions (Artis and Ostry (1986, p. 21)). In addition to coordination proper, it also extends to looser forms of cooperation, including consultation among countries, information interchange, and international surveillance.

Convergence refers to developments in which the levels or rates of growth of economic variables move closer to each other over time.3 These developments may refer to either policies or economic performance. Convergence in economic performance, however, may require divergence in underlying economic policies. Furthermore, convergence in either policies or performance need not require coordinated or cooperative policies, nor may coordination of policies necessarily imply convergence of those policies.

Harmonization of economic policies refers to the application of national regulations with a view to achieving greater uniformity in economic structure across countries (Steinherr (1984, p. 73)). The term has, in general, been used in a fairly restrictive sense to refer to microeconomic policies; for example, the adoption of specific taxation and tariff policies within the European Communities (EC). Because the concern of the present study is with macroeconomic policy issues, the concept of harmonization thus defined is not extensively discussed.

Alternative Forms of Cooperation

International economic cooperation can take different forms, including the exchange of information among countries, policy coordination, and the adoption of rules governing the operation of the international monetary system. The simplest form of cooperation is the exchange of information among countries. Systematic collection of information about the policy intentions of, and recent developments in, other countries is vital to effective policy coordination. This form of international economic cooperation imposes the least constraint on national policymaking. Collection of relevant information about developments in important macroeconomic variables—such as the balance of payments and growth and inflation in the short term and medium term—may act to identify possible global inconsistencies and the need for policy coordination. In recognition of the importance of further strengthening this form of cooperation, the Fund has recently been requested by the Interim Committee to consider the use of indicators of policy actions and economic performance in its surveillance over members’ policies.

Policy coordination itself can be divided into two types: ad hoc or episodic coordination 4 that is characterized by discussion among the interested parties, with action undertaken only when agreement is reached; and institutionalized coordination, or centralized decision making, by which decisions are made either collectively by member countries (for example, in the European Monetary System, EMS), or by an institution acting on behalf of its membership. To date most examples of coordination have been episodic, although international institutions contribute to the information exchange and surveillance that are important forms of international cooperation. Another useful distinction is between those coordinated policy decisions that involve short-term discretionary demand management (“fine-tuning”), and those that are taken with regard to their medium-term consequences. International coordination of policies for short-run stabilization purposes is subject to the same criticism as domestic fine-tuning: that in the absence of detailed information about the effects of policies and their credibility, such coordination may do more harm than good. For instance, policies typically operate with a lag, but the length of the lags is not precisely known; in these circumstances economic stimulus applied, for example, to counteract a recession may occur at an inappropriate time.

Plan of the Paper

The paper first presents the reasons that cooperation and coordination may be desirable. It is argued in Section I that uncertainty about other countries’ policies and their effects can be reduced by sharing information. Pooling information on the stance of policies may make evident that individual policies are likely to be unsustainable in the future and, further, that policies are incompatible from a global perspective. The form that this global inconsistency may take depends on the exchange rate regime. Section I also considers necessary conditions for policy coordination—in the sense of joint decision making by a group of countries—to be desirable.

The operation of the present international monetary system, which is characterized by a wide diversity of exchange rate agreements among countries, is described in Section II. Examples of recent experience with policy coordination are provided, including developments within the EMS and the meetings of members of the Group of Seven (G-7: the United States, Japan, Federal Republic of Germany, United Kingdom, France, Italy, and Canada).

In Section III, the costs and benefits of coordination are discussed in the context of the present international monetary system and from the perspective of recent theoretical and empirical studies. The broad conclusions of this discussion are that, although the gains from coordination are clear when the effects of policies are known, they are less likely to be realized in a context of uncertainty about the way the economic system operates. There are also obstacles to the achievement of coordination and incentives not to implement agreed policies. Nevertheless, these obstacles make international economic cooperation—in particular, monitoring the transmission of economic policies and sharing information—all the more essential.

Section IV considers ways in which international economic cooperation can be made more effective and reviews various recent proposals in that light. Because appropriate policies depend on the form of the international monetary system, on the nature of transmission mechanisms, and on the kinds of shocks affecting the world economy, any attempt at enhanced coordination of policies must be implemented in a flexible fashion. Flexibility would suggest that schemes for coordination that rely on a single indicator—for example, the exchange rate—are unlikely to be optimal, primarily because they do not exploit all the information that is potentially available to policymakers, and may thereby give misleading signals for appropriate policy adjustment. Nevertheless, a clear analytical framework for devising and interpreting a set of key indicators is indispensable in advancing knowledge about the effects of policies and, thereby, in providing information on both the policy choices open to national governments and the scope for coordinating their policy actions. Finally, Section V presents brief conclusions.

I. Sources of the Perceived Need for Cooperation and Coordination

It is the interdependence of national economies that makes international economic cooperation desirable in some circumstances. If a country were completely independent of its neighbors, then its policy choices would obviously not be complicated by disturbances emanating from abroad or by the spillover effects of policies taken by other countries. In these circumstances there would be no need for obtaining information about developments abroad or for coordination of economic policies. Needless to say, this is not a realistic description of the situation facing any country today; in recent years goods and financial markets have become much more integrated, tending to increase economic interdependence (Cooper (1986a)).

Interdependence among countries takes a number of forms. Structural interdependence (Cooper (1986a, p. 292)) implies that events in one country strongly influence those in another. The main channel through which this influence occurs is trade in goods and financial assets; in general, greater integration of goods and asset markets between countries is likely to lead to greater interdependence of the economies concerned. Correlation of the exogenous disturbances facing two economies may increase their linkage for a given level of structural interdependence (Cooper (1986a)). For instance, because of its magnitude and global effect, the oil price shock in 1973–74 produced essentially the same stagflationary effects in all industrial economies and increased interest in economic cooperation, in particular in sharing the large current account deficits of those countries with respect to OPEC (Organization of Petroleum Exporting Countries). It may also be the case that there is interdependence of the objectives of policy. For instance, countries may share a concern for world inflation (not just inflation in their own country) or for economic growth in developing countries.

The global economy can be considered a closed economy, with the implication that there are “adding-up” constraints that apply to variables across countries (Hamada (1979)). For example, trade balances and current account balances have to sum to zero (apart from any errors of measurement), and the change in reserves—or the overall balance of payments—when summed over all countries must equal the global supply of reserves. As a result of these constraints, if a country takes action to change its balance of payments, then this action will have repercussions for at least one other country. These “zero-sum” constraints on such national policy objectives as current account balances and holdings of international reserves give rise to the likelihood of international inconsistencies in policy targets. These inconsistencies, in turn, imply a potential need for cooperation and coordination among countries.

Interdependence implies that it is important to take into account the policy settings of other countries, and to understand the way in which economic policies and exogenous shocks are transmitted between countries. Thus there is a crucial role for sharing information about international developments and for explaining policy intentions to other governments. Given the rudimentary understanding of the precise nature of economic relationships, including possible lags in their operation, economic cooperation in extending economic knowledge about both the structure of the economy and the current and future values of relevant macroeconomic variables is likely to yield a high payoff. It is probably fair to argue that, because of the complexity of the international economy, policymakers at present do not fully understand developments abroad—let alone incorporate them into their policy choices.5 International economic cooperation in the form of information sharing helps those policymakers to understand better how the effects of their policies are modified by events abroad—including the feedback effects of their own policies. The case for cooperation does not, however, rest on the argument that those feedback effects are large. In the context of uncertainty about the state of the world economy, even the limiting case of a country that is too small to have any influence on other countries—and hence is free of feedback of its own policy actions from abroad—would still want to engage in international economic cooperation of this form. For a small country, such cooperation might be all the more necessary.

As noted earlier, it is useful to distinguish between the need for cooperation in the form of information exchange and economic policy coordination. The exchange of information, and a framework for interpreting that information, is an essential first step in cooperation. Achievement of this step would imply that governments, even when acting independently, could choose policies that would be less likely to lead to unexpected international feedback effects, thereby permitting them to achieve results closer to their objectives. In some circumstances, however, cooperation must also involve a further stage: the adjustment of policy goals in each country to make them internationally consistent and compatible (Cooper (1986a, p. 316)).6 This need is most obvious when countries target the same variable—for instance, if two countries each target their bilateral exchange rate or bilateral current account balance.

The source of possible incompatibilities has been called the “nth-country problem,” which manifests itself in different ways under different exchange rate regimes. Under a key-currency system with fixed exchange rates, central banks agree to intervene in the foreign exchange market to maintain a parity rate against the key currency, and balance of payments adjustment is achieved through the feedback of reserve flows onto the domestic money supply (in the absence of sterilization). The induced change in the money supply will have effects on output and prices that will tend to correct the payments disequilibrium. Typically, countries will not be indifferent to the levels of their reserves. If all countries target the level of international reserves, and the sum of reserve targets exceeds the world reserve stock, then policies have a global deflationary bias; if they fall short of the supply, then the world economy will experience inflationary pressures (Hamada (1974)). If the key-currency country (the n th country) is willing to act passively as a residual and does not adopt a (net) reserve target, it ensures that the global supply of reserves can eventually be distributed according to the preferences of the other n − 1 countries. The n th country also has the freedom to set its own money supply, since other countries ensure the maintenance of nominal parities. Under the Bretton Woods system, it was the United States that had the role of the nth country. The system broke down when other countries were unwilling to accept the inflationary consequences of U.S. money supply growth and when the United States became increasingly unwilling to accept passively developments affecting its own balance of payments.

In principle, a regime in which exchange rates float freely eliminates the need for official reserves and allows each country to choose its money supply and rate of inflation. In practice, of course, no country is indifferent to its current account balance, its real effective exchange rate, or the level of its reserve holdings. If countries have targets for these variables, there must either be an nth country that does not have an explicit target, or there must be some way of making the targets globally compatible. This is the fundamental consideration that has given rise to the impetus for policy coordination in the present international monetary system, with its high degree of exchange rate flexibility. Although the present system allows member countries almost complete freedom of choice about their exchange rate arrangements, a stable system of exchange rates remains of international concern and the object of Fund surveillance.7 In the absence of cooperation to achieve this objective, market mechanisms may establish compatible but nonoptimal outcomes, and persistence in trying to achieve those goals may destabilize the international economy.

The international incompatibility of policies is closely related to their unsustanabity. Policies are unsustainable, however, for their ultimate consequences—both domestically and in the world economy—even if they are not currently incompatible with other countries’ policies. An example is a fiscal policy that involves large government deficits, a persistent increase in the ratio of government debt to gross national product (GNP), and large current account deficits. Such a policy may not be immediately incompatible with other countries’ targets—indeed, the current account deficit may be viewed favorably abroad because of its stimulus to output in those countries, and capital inflows may occur as international asset holders take advantage of attractive returns—but the sustainability of such a policy is a valid concern to the authorities because of the need to identify problems that may arise in the future. International economic cooperation, as distinct from policy coordination, may play a useful role in drawing attention to the unsustainability of policies, which may in some cases be more evident from an international than from a domestic perspective.

A demanding form of cooperation is economic policy coordination; this form has received much attention recently in the academic literature (see, for example, Buiter and Marston (1985), Cooper (1986a), Corden (1985), and Frenkel (1986)). The theoretical discussions have usually assumed that the economic structure of each country involved in the effort of policy coordination is known with certainty by all parties, as are the policy actions of all countries. In this context gains from coordination, in the sense of the improvement of each country’s welfare relative to the outcome with independent policy setting by each country, require not only interdependence of national economies and the inability of governments to achieve their objectives perfectly (in other words, they have fewer independent instruments than targets), but also that feedback effects from abroad change the trade-off between policy goals (Gavin (1986)). For example, if these feedback effects fail to change a country’s position along its short-run Phillips curve, there will be no gains from policy coordination. Whether these conditions are satisfied is the subject of controversy; detailed knowledge of the economic system is essential to resolve the issue of whether coordinated policies should be different from policies that governments would have chosen independently, and in what direction coordinated policies would be different.

Coordination of economic policies may in some cases be desirable, but uncertainties about the precise goals of policymakers, the nature of economic transmission mechanisms, and the kinds of shocks affecting national economies make it difficult to generalize with regard to the improvement that coordination would bring. Because the form of cooperation or coordination depends on the way the international monetary system functions, a discussion of experience with the generalized flexibility of exchange rates as it has operated since 1973 is in order.

II. Operation of the Present International Monetary System and Experience with Coordination

The present international monetary system is characterized by a wide diversity of exchange rate arrangements among countries. These arrangements include some form of fixed exchange rates achieved by pegging against individual currencies or a currency basket (adopted by many developing countries), limited flexibility of exchange rates against a single currency or group of currencies (as under the exchange rate mechanism of the EMS), and greater exchange rate flexibility (practiced by some industrial countries, notably the United States, Japan, Canada, and the United Kingdom, and by some developing countries). Although the actual number of countries that choose to operate under a fixed or adjustable peg system far exceeds those countries with floating currencies, the behavior of the latter group dominates the functioning of the international monetary system by virtue of its disproportionately large share of world trade.8

The existence of interdependencies makes it possible that the smooth functioning of the present “hybrid” system of exchange arrangements would be enhanced by some coordination of macroeconomic policies. The appropriate form that coordination might assume for different groups of countries depends, however, on the nature of the prevailing exchange rate arrangement. At the same time, the diversity of the exchange rate system increases the difficulties of defining rules or cooperative policies that are appropriate within and between groups of countries. The complexity of the present system also means that any policy lessons based on an analysis of the “pure” cases of fixed and flexible exchange rates may require considerable modification in light of the actual experience of countries within the current international system (McKibbin and Sachs (1986) and McKinnon (1984)).

Lessons from Experience of the Seven Major Industrial Countries

The post-Bretton Woods experience of the main industrial countries may be characterized as independent floating, although the considerable intervention that has taken place at certain periods blurs any sharp distinction based solely on an analysis of the pure regimes of fixed and flexible exchange rates.9 Four criteria might serve as a basis for evaluating the experience of these countries under the present exchange rate system (International Monetary Fund (1984b)): (1) does the system help or hinder macroeconomic policy in the pursuit of domestic objectives? (2) how effective is the system in promoting external adjustment? (3) how does the system affect resource allocation in the world economy? (4) how flexible is the system to changes in the world economic environment? Much of the recent interest in coordination has stemmed from a concern with the effectiveness of the present system, and from a perception that there is excessive exchange rate volatility, misalignment of the currencies of the main industrial countries, and continuing payments imbalances, especially of the United States, the Federal Republic of Germany, and Japan. Figure 1 presents the evolution of current account balances and real effective exchange rates for the major industrial countries since the start of generalized floating. It indicates that there have been large swings in current account balances as ratios to GNP, especially since 1982 in the case of the United States and Japan, and correspondingly large swings in real effective exchange rates.

Figure 1.Current Account Balances and Real Effective Exchange Rates of Major Industrial Countries, First Quarter of 1973 Through End-1987


Sources: International Monetary Fund (1987 and various issues) and Fund staff estimates.

a Relative normalized unit labor costs adjusted for exchange rate changes, calculated using export weights. (Left-hand scale.)

b Seasonally adjusted, including official transfers. (Right-hand scale.)

c Current account data for Japan and Italy extend through the third and second quarters, respectively, of 1987.

As several observers have noted, there are considerable methodological problems in determining the extent of payments imbalances and identifying the role played by the exchange rate system.10 Ideally, what is sought is a comparison of underlying payments imbalances (actual payments balances adjusted for factors such as the effects of lagged exchange rates and cyclical effects) with equilibrium payments balances.11 An additional problem arises in attempting to isolate the effect of the exchange rate regime from the period itself. For example, the immediate post-Bretton Woods period was characterized by large real disturbances leading to a marked deterioration in the terms of trade of the main industrial countries, but this was not the result of the exchange rate regime. Nevertheless, the current account balance is used here as a crude measure of payments imbalances, and Table 1 compares the level and persistence of payments imbalances since 1973 with measures for the preceding decade.12

Table 1.Current Account Balances as a Percentage of Gross National Product (GNP): Major Industrial Countries (Group of Seven), 1963–72 and 1973–86
Mean aDeviationCorrelationb
United States0.30-0.700.461.401.02*1.00*
Germany, Fed. Rep.of0.480.981.091.530.59*0.78*
United Kingdom0.30-*
Unweighted mean0.690.640.951.450.54d0.66d
Source: International Monetary fund (1987 and various issues).Note: The current account includes goods, services, and all current transfers, both official and private.

Country means take into account the sign of current account imbalances. In contrast, group means are based on absolute values of country means.

The statistic reported is the estimated coefficient on the lagged dependent variable in the first-order autoregression equation; an asterisk indicates statistical significance at the 95 percent level.

1967–72 only.

Excludes France.

Source: International Monetary fund (1987 and various issues).Note: The current account includes goods, services, and all current transfers, both official and private.

Country means take into account the sign of current account imbalances. In contrast, group means are based on absolute values of country means.

The statistic reported is the estimated coefficient on the lagged dependent variable in the first-order autoregression equation; an asterisk indicates statistical significance at the 95 percent level.

1967–72 only.

Excludes France.

Table 1 shows that, although some individual countries—notably the United States, Canada, and the Federal Republic of Germany—have experienced higher average and more variable current account imbalances under flexible exchange rates than in the last decade of Bretton Woods, this was not true for an unweighted average of the G-7 countries. On this basis, these countries experienced smaller payments imbalances under flexible exchange rates, although their variability over time increased. As regards the persistence of current account imbalances, the evidence indicates that it has increased under exchange rate flexibility. In contrast, early proponents of floating exchange rates argued that flexibility would ensure prompt adjustment of external imbalances. For the United States, Japan, and Germany, imbalances have been large and persistent under flexible rates.

Recent Attempts at Economic Policy Coordination Among Major Industrial Countries

The post-Bretton Woods experience of independent or joint floating by the major industrial countries has also been characterized by both synchronization of domestic policies that have at times ignored the amplification of spillover effects from other countries and episodic attempts at cooperation and coordination. Since 1975 an important new instrument for achieving international coordination has arisen—the annual Economic Summit of the G-7 countries.

Pressures for coordination are likely to arise when the economic system is perceived as not working, and the Economic Summit was a response to a need for international cooperation following the breakdown of the Bretton Woods system and the 1973 oil shock. Major industrial countries underwent a marked deterioration in economic performance after 1973, with sizable increases in both inflation and unemployment. This was accompanied by a sharp fall in productivity and output growth rates, both of which declined in 1973–75 to less than half of their respective averages for the previous decade.

The distinctive feature of the Economic Summits is that they are limited to a small group of countries (currently the G-7) that carry most of the weight in international economic decision making. One important advantage of limiting the number of participating countries is that it reduces the number of conflicts and costs of negotiating. Unlike more permanent institutions, summits do not impose binding agreements on countries, and the subjects considered may alter each year. This flexibility has proven to be both an advantage (in the face of a changing economic environment) and a disadvantage on occasions, as illustrated in the discussion below.

In assessing the record of the G-7 Summits, it is useful to distinguish between the procedural and substantive achievements that have been made. On a procedural level, the Summits have established an increased awareness of policy interactions, a recognition of the role of exchange rate factors in macroeconomic policy formulation, and the need for mutually consistent medium-term strategies. The significance of this accomplishment should not be understated. At a substantive level, the policy initiatives agreed to at the G-7 Summits have gone less far. The earlier Summits placed considerable emphasis on the importance of each country “putting its own house in order.” The outcome was frequently a synchronization of domestic policies, as illustrated in the period following the 1979–80 oil shock when the main industrial countries independently implemented policies of monetary restraint. The synchronized tightening in this case led to historically high real interest rates and a generalized contraction in economic activity, with adverse consequences for economic growth in developing countries and for their capacity to service external indebtedness.

One important exception to this domestic focus was the Bonn Summit of 1978, which achieved agreement on a coordinated package of macroeconomic policies (as interpreted in the broad definition of this paper).

Specifically, Germany agreed to measures that would increase its 1978–79 budget deficit by about 1 percent of gross domestic product (GDP), Japan promised additional public expenditure, and the United States implemented an energy package to reduce oil imports. The form of macroeconomic coordination implemented was, to a large extent, a modified version of the “locomotive” model that had dominated discussion at the London Summit a year earlier. It was believed that the larger industrial countries should undertake more expansionary policies than otherwise to foster the conditions for an export-led world recovery. Whatever the economic merits of the Bonn measures, the success in reaching a coordinated agreement may be attributed to several factors (Putnam and Bayne (1984) and Artts and Ostry (1986)): (1) a temporary synchronization of political ideologies; (2) an integration of the Bonn Summit agenda with lower-level discussions under the auspices of international organizations: and (3) a meshing of domestic and international issues in the sense that domestic advocates of internationally desired policies were able to use the Summit to shift the internal balance of power in their favor.

An assessment of the success of the economic measures implemented at Bonn is difficult and controversial in view of the impact of the second oil shock in 1979. By the end of 1978, recovery in economic activity was under way, but it was then halted by the 1979–80 oil price increases. The sharp economic downturn that followed was also marked by historically high nominal and real interest rates that accompanied restrictive monetary policies. In retrospect, it can probably be said that the Bonn measures placed insufficient emphasis on the medium-term consequences of fiscal expansion and failed to provide sufficient flexibility for anti-inflationary monetary policies that were subsequently made necessary by the second oil shock. As a result, the expansionary measures decided at Bonn were soon revised, and the Bonn Summit is widely considered to be an example of the pitfalls of international fine-tuning.

The coordinated reductions in the discount rates of major industrial countries achieved in March and April 1986 may be interpreted as an example of successful coordination of monetary policies when there is a common perception of the goal to be attained—in this case, the need to lower nominal (and real) interest rates in the face of slow growth of output. The discount rate reductions were also justified by the desire to avoid further exchange rate changes, in light of the substantial depreciation of the dollar that had already taken place. The commitment to policy coordination was further reaffirmed at the Economic Summit in Tokyo in May 1986, and again with the Louvre Agreement in February 1987 and at the June 1987 Venice Summit, in continuation of the process that began with the Plaza Accord in September 1985. The mixed success in achieving similar coordinated interest rate reductions among the main industrial countries since that time, however, also highlights an important limitation of the Economic Summit as an instrument for achieving international coordination. It may prove difficult to achieve continuing agreements among countries through ad hoc policy coordination in the face of changing circumstances, differing interpretations of the nature of agreements, and use of different macroeconometric models.

The main lessons that might be drawn from the recent experience of the major industrial countries with international cooperation are not only the difficulty of reaching a coordinated agreement but also the problem of determining the appropriate response in the form of synchronized or nonsynchronized monetary or fiscal policies. Reaching international agreement is often complicated by a lack of domestic consensus and, in some cases, by the absence of political will to carry out policies that are judged to be desirable. But it is often unclear what the appropriate policies should be, and this difficulty reflects a lack of information about how the economic system works, especially within a changing economic environment. (This issue is considered in more detail in Section III below.)

Lessons from Experience of the European Monetary System

Useful lessons for international policy coordination may also be drawn from the experience of the group of EC member countries whose currencies participate in the exchange rate mechanism (ERM) of the EMS.13 From an analytical perspective, the ERM operates as an adjustable peg system with intervention by central banks to maintain nominal exchange rates within, or at the margins around, bilateral central rates of participating countries (see Ungerer and others (1986)). Unlike the Bretton Woods system, in which a parity change in a situation of “fundamental disequilibrium” could be initiated unilaterally by member countries, the direction and magnitude of realignments within the EMS are determined by collective decision making.

The goals of the EMS have a critical bearing on any evaluation of its performance and the extent to which this can be attributed to policy coordination. The stated objective of the EMS is to create a zone of monetary and exchange rate stability, since this stability is viewed as a prerequisite to achieving convergence in economic performance. The primary focus on monetary stability has meant that the control of inflation has formed a cohesive primary objective for member countries whose currencies participate in the exchange rate arrangement.14 The agreement to place a high weight on reducing inflation has also served to weaken any threats to stability that might otherwise arise from conflict between internal and external balance, as well is any inconsistency between inflation targets desired by smaller members and larger countries. At the same time, the critical importance of inflation reduction in uniting ERM partner countries suggests a possible source of future instability and policy conflicts if the priority were to shift in some countries toward reduction of unemployment.

The goal of convergence of economic performance within the EMS, resulting from a general desire to reduce inflation rates from their high levels after the oil price shock of 1979, has required the achievement of some degree of convergence of economic policies. It is not the case, however, that convergence of policies can be equated with coordination of policies.15 Most writers have broadly identified coordinated policies of EKM countries with convergent policies while recognizing the possible inconsistencies that might arise between short-term and long-term goals. Although the convergence toward monetary restraint in ERM member countries (measured by monetary and domestic credit aggregate growth rates) has been well documented, it is also apparent that this experience is not unique to this group of countries (Ungerer and others (1986) and Artis (1987)), A similar movement toward greater monetary restraint has been exercised in other industrial countries that have operated under more flexible exchange rate arrangements over the same period. Fiscal policies, in contrast, were not directed toward the exchange rate, and showed increasing divergence for ERM (and major non-ERM industrial countries) over the period 1979–86, although the EMS may have placed constraints on fiscal policies of particular countries in specific periods (for example, the reversal of the fiscal expansion undertaken by France in 1981–82).

Studies of the exchange rate performance of the ERM (Ungerer and others (1986), Artis (1987), and Rogoff (1985b)) show a reduction in the volatility of both actual nominal and real exchange rates among member countries and of the unanticipated component of those rates. If it were the case that greater predictability of exchange rates could be attributed to a change in private speculative behavior in response to a credible commitment to a set of central rates among ERM countries, then we might also expect to observe a corresponding reduction in the mean and variability of nominal and real interest rate differentials among ERM countries over the same period. Recent analysis by Rogoff (1985b) suggests that average unanticipated real interest rate differentials for both France and Italy against Germany showed a rise over 1979–86. This finding is problematical insofar as it suggests that the observed exchange rate stability may derive from both the imposition of capital controls and from a willingness to use monetary instruments to resist exchange rate movements.

Measures of convergence of economic performance of ERM member countries, including inflation, unemployment, and current account imbalances, are summarized in Table 2, These statistics show a mixed record. EMS countries participating in the ERM have indeed experienced a convergence in price performance as measured by a fall in both average inflation rates and the dispersion of these rates around mean levels, although this fall is concentrated in the most recent 1982–86 period. Again, this experience is paralleled by that of the group of nonmember countries. In contrast, indicators of unemployment and current account imbalances show an increased divergence for both ERM countries and G-7 countries not included in the exchange rate arrangement. Although the considerable increase in average unemployment rates for ERM members may be attributed to real factors and also to the revealed preference of policymakers for reducing inflation, it may also reflect unexploited opportunities for coordination of other policies, especially fiscal policy and microeconomic measures.

Table 2.Indicators of Convergence: Averages for Inflation, Unemployment and Current Account Balances in Exchange-Rate-Mechanism (ERM) Participants in European Monetary System and in Non-ERM Members of G-7. 1974–86
Consumer Price IndicesUnemployment RatesCurrent Account Balances
Annual change in percentPercentPercent of GNP
ERM countries
Fed. Rep. of4.723.591.903.606.348.091.350.712.13
Weighted mean a9.337.795.304.768.279.821.852.572.24
Weighted standard
G-7 non-ERM countries
United Kingdom16.188.874.774.389.1011.18-1.130.880.53
United States8.006.783.246.997.707.820.02-1.31-2.61
Weighted meana9.516.193.075.666.717.050.911.031.60
Weighted standard
Source: International Monetary Fund (1987 and various issues) and World Economic Outlook data bank.

Weighted by each country’s GNP. For current account balances, the mean and standard deviation are unweighted: country means take into account the sign of current account imbalances, whereas group means are based on absolute country means.

Source: International Monetary Fund (1987 and various issues) and World Economic Outlook data bank.

Weighted by each country’s GNP. For current account balances, the mean and standard deviation are unweighted: country means take into account the sign of current account imbalances, whereas group means are based on absolute country means.

The short period of existence of the EMS and its continuing evolution mean that any policy lessons that might be drawn must be qualified ones. Furthermore, the EMS experience may be viewed as unique in that participating countries have a high degree of political commitment, similar economic structures, and, for some countries at least, capital controls. Nevertheless, some comments can be made based on the earlier discussion and the evidence presented in Table 2. First, although the discipline of fixed but adjustable exchange rates may have served as the chief mechanism for achieving convergence in inflation for ERM countries, it is clearly not a necessary instrument for monetary discipline. Second, it appears that the ERM is not a sufficient instrument for achieving convergence in other measures of performance, in particular unemployment and the reduction of current account imbalances, unless it is also accompanied by coordination of both fiscal and monetary policies.

III. Cost and Benefits of Coordination in the Present Exchange Rate System: Lessons from the Literature

The previous section sketched some of the features of the present international monetary system and discussed some attempts at policy coordination. In this section the theoretical and empirical literature on transmission mechanisms is surveyed, given the importance of these mechanisms for coordination. The obstacles to achieving coordination are then discussed, as are factors that suggest that coordination may not always be a good thing. Finally, the relatively few empirical studies on policy coordination are surveyed.

Transmission Mechanisms Under Fixed and Flexible Exchange Rates

The particular exchange rate system that is adopted may also affect the nature of the interdependence among countries—in particular, the transmission of economic policies. If countries were not affected by the policies of others, then there would be no externalities to be eliminated by coordination. The hopes of the many advocates of exchange rate flexibility were that it would bring about a greater degree of independence among countries; however, experience with the currencies of the major industrial countries since the advent of generalized floating in 1973 has been that interdependence is not eliminated by exchange rate flexibility. On the contrary, with increased capital mobility, floating exchange rates have made countries more sensitive to certain external shocks.

There is a voluminous literature on the transmission of the effects of economic policies under fixed and flexible exchange rates.16Mundell (1962) and Fleming (1962) analyzed the transmission of policies from the perspective of high capital mobility and fixed wages and prices. They further assumed that, given inelastic exchange rate expectations, high capital mobility would equate interest rates at home and abroad. Under fixed exchange rates, monetary or fiscal expansion in a large country is likely to be positively transmitted to other countries by increasing aggregate demand at home and also imports from abroad. In contrast, under flexible exchange rates a monetary expansion in a large country in the Mundell-Fleming model leads to an increase in output at home but a decline in output abroad. A small country cannot affect its interest rate, but a monetary expansion will stimulate output and depreciate the exchange rate. As regards fiscal expansion, it is completely ineffective as a tool for increasing domestic output in the case of a small country, since the exchange rate appreciates sufficiently to crowd out the increase of demand at home. Otherwise, the conditions for monetary equilibrium would not be satisfied. Fiscal expansion in a large country will, however, have some effect in stimulating output at home and abroad, and this stimulus will be associated with a rise in the world rate of interest. Thus, fiscal expansion is transmitted positively under flexible exchange rates in the Mundell-Fleming model.

Two important elaborations on that model for flexible exchange rates involve accounting for exchange rate expectations in the context of sticky goods prices and modeling aggregate supply. Both of these elaborations have the effect of weakening the clear conclusions about the direction of links among countries under flexible exchange rates. Dornbusch (1976) has allowed for sticky—not fixed—prices and for exchange rate expectations that correctly anticipate future movements. In this model, interest rates need not be equal at home and abroad, and an increase in government expenditure can cause an expansion in output, even in a small country. Moreover, monetary expansion need not be transmitted negatively to the foreign country, and monetary shocks cause overshooting of the nominal exchange rate in the short run. Models of aggregate supply 17 have exploited the distinction between the real wage calculated from the point of view of the firm and the real wage calculated from the point of view of the worker, or consumer. When effects of the terms of trade on the consumption wage are taken into account, a fiscal expansion at home is less likely to cause an expansion of output abroad, and a monetary expansion at home is more likely to be associated with a rise in output abroad.

Obstacles to Coordination

Knowledge of how the economic system operates is far from perfect, as has been illustrated above in the discussion on transmission mechanisms and the experience of countries under the present international monetary system. First, disagreements about the effects of a particular policy can also at times be great both among different arms of the same government and between governments. The lack of consensus—indeed conflicts in views—about a model of the economic system may well make agreement impossible; or, if governments do manage to reach agreement, then there is no guarantee that welfare will actually be improved. Second, there also are costs to negotiating and policing agreements when there are incentives to renege on those agreements and a reluctance of governments to limit their freedom of maneuver. These considerations may seriously limit the scope for international policy coordination. Third, instead of viewing governments as maximizing the welfare of their citizens, it may be more realistic to consider coordination in a context in which the aims of the government and the private sector may differ, as might be the case, for instance, if political leaders’ planning horizons did not extend beyond the next election. A consequence of this viewpoint is that coordination among national governments will not necessarily improve the welfare of the private sectors in those countries, and hence it may not be desirable.

Lack of Consensus About the Functioning of the Economic System

A reason for inability to achieve policy coordination is disagreement about the way the economic system functions; in particular, about the strength and even the direction of the transmission mechanisms for economic policies. As discussed above, existing models give conflicting signals about the direction of the transmission mechanisms. Cooper (1986b) has argued that international cooperation in the area of disease prevention took the good part of a century, despite obvious advantages to all countries, because of lack of firm scientific evidence on the nature of disease transmission. Once a scientific consensus emerged, progress was rapid. By analogy, success in economic cooperation is likely to be greatest where the knowledge of the effects of policies is most developed. An example of successful cooperation may be the harmonization of the value-added tax and the standardization of packaging and labeling regulations within the EC. In contrast, there is much disagreement about the effects of macroeconomic policies on policy objectives.

Disagreement among domestic policymakers may also be an obstacle to international coordination. In most countries domestic objectives reflect the outcome of considerable bargaining among interest groups, which may leave little scope for compromise at an international level (Polak (1981)). The view is cited in McKibbin (1985), however, that international agreements have been achieved in recent years precisely when there have been substantial disagreements between national authorities, for example between monetary and fiscal authorities, and there consequently has been some scope for coalitions of officials with similar objectives to form across countries.

Costs of Negotiating and Policing Agreements

There may be advantages to some countries from a noncooperative solution if they can impose their policies on others (Steinherr (1984)). To use the terminology of oligopoly theory, a Stackelberg solution, in which one country acts as leader and another as a follower, may benefit the former in some cases. 18 An example may be the operation of a pegged exchange rate system in the 1950s and early 1960s that allowed the United States to implement its monetary and fiscal policies with little immediate concern for their repercussions for the U.S. balance of payments, since other countries played the role of follower and ensured that their own policies were consistent with their dollar pegs. This situation, which gave the United States an extra degree of freedom, was tolerated by other countries only so long as U.S. policies were perceived as neither too inflationary nor too deflationary. Inflationary pressures in the United States eventually led to the breakdown of the system.

In other circumstances, countries may benefit from being followers; this is also termed the “free-rider problem,” which can occur if there is an international public good from which all countries benefit, but to which each would like to minimize its contribution. An example that is sometimes cited is the attempt to stimulate aggregate demand by increased government expenditure. If fiscal stimulus is transmitted positively to other countries, then all will experience an increase in demand; however, each country individually may not want to carry out fiscal stimulus because of unfavorable balance of payments consequences. The optimal outcome from the point of view of a country suffering from deficient demand and pressure on its balance of payments would be to have other countries stimulate demand. This was the situation of several industrial countries during the 1970s, after the stagflationary oil price shock; hence the interest in “locomotive” and “convoy” theories to put pressure on Germany and Japan to expand domestic demand (Artis and Ostry (1986)). 19

Another issue that has received much attention in the academic literature on optimal policy choice in a closed economy is “time inconsistency.” starting with the seminal work of Kydland and Prescott (1977). Time inconsistency refers to a problem associated with the possibility that authorities will find it advantageous to change their plans in the future. Suppose that a government, maximizing a given objective function, chooses an optimal monetary policy designed to bring down the rate of inflation; the optimality of this policy is predicated on the presumption that a restrictive policy in the future, if anticipated today, will have the effect of lowering expectations of inflation, and thereby will have an immediate effect on inflation. If this occurs, then the government will have obtained the benefits of the tight policy, without actually having implemented it; when the time comes to incur the unemployment cost of the policy, the government will therefore not have the same incentive to pursue its restrictive policy. Because private agents also are aware of this problem, they will assume that the government will not carry out its announced policy (which will no longer be optimal), and hence will not lower their inflation expectations.

But, if policymakers can acquire a reputation for not reneging on agreements, they may reap the benefits of optimal, but time-inconsistent policies. In this case they can credibly commit themselves to future policy actions, even though at some point it might seem optimal to renege. The same considerations apply to the possibility of reaching optimal policies through cooperation among national governments. In some circumstances, it may seem attractive for governments to negotiate agreements—for instance, to stimulate aggregate demand—but to fail to “carry out their end of the bargain” although benefiting from the actions of the other parties to the agreement. The incentive toward this type of behavior, if perceived to be pervasive, might make cooperative behavior impossible unless there are penalties attached to noncompliance.

Although this obstacle to cooperative behavior at the international level has a certain plausibility, there are several considerations that limit its import. As Canzoneri and Henderson (1988, forthcoming, Chapter 4) have argued, considerations of reputation may allow a cooperative equilibrium to be achieved, even when incentives exist to renege on agreements. Using techniques of game theory, they show that an equilibrium with optimal policies can be achieved provided that there are credible threats of retaliation in later periods if a party to an agreement behaves noncoopcratively. It thus appears that the absence of international cooperation is more likely to be due to other considerations—for instance, the uncertainty about the effects of policies discussed above—and that cooperation will be more successful if there are clear “performance criteria” and it is easy to verify that a party is holding up its side of the bargain. The monitoring aspect is easier to achieve if there is regular exchange of information and contact within established institutions, such as the EC. A related issue is that cooperation in several fields may allow bargains to be struck such that a gain in one area is traded off against concessions in another; thus agreements may be easier if policy packages are concluded on a regular basis (Steinherr (1984)).

Coordination among governments is costly in terms of the negotiating process and time lags in reaching agreement. Economic Summits and negotiations among lesser officials may involve a considerable diversion of attention away from domestic economic and political issues (Artis and Ostry (1986)). Furthermore, agreements, even if achieved, are likely to involve substantial further efforts of implementation because there are often incentives for cheating that must be minimized—possibly through monitoring, cajoling by the other participants of the agreement, or threats of retaliation.

The costs of coordination are likely to rise with the number of participants, so that limited cooperation, whether among the larger industrial countries or within regional groups, is more likely than cooperation on a global basis. The effectiveness of Economic Summits might also be enhanced by representation from international institutions, although in practice the spring Interim Committee and OECD meetings indirectly contribute to the preparations of a Summit. Such institutions may contribute significantly to the negotiating process by their global focus as well as by representing the interests of developing and smaller industrial countries. Nevertheless, the economist’s Pareto-optimal cooperative policy is unlikely to emerge. Given the policing costs of ad hoc agreements and the efforts needed to consider issues of policy on an ongoing and time-consuming basis, changes in the rules governing the international monetary system would seem obviously preferable, provided it were possible to codify rules that proscribed recourse to “bad” policies. Indeed, Polak (1981) argues that explicit coordination should be considered a second best; where it is clear what rules should guide policies, then these rules should be subject to administration and policing by international organizations rather than be subject to continuous renegotiation. In summary, although the theoretical case in favor of policy coordination may be strong, there are real-world problems that may limit its practical possibilities.

Is Coordination Necessarily a Good Thing?

The discussion until now has been based on the presumption that, although there may be obstacles to achieving it, coordination is necessarily desirable because it permits moving closer to the Pareto-optimal solution. More fundamentally, it may not be true that coordination improves global welfare (Rogoff (1985a)). It could be argued that, because coordination is voluntary, it surely would not take place if welfare was expected to be reduced. But coordination involves only a subset of actors in the economy, and consequently there is no guarantee that it leads to a better solution than independent behavior, even leaving aside uncertainty about the effects of policies.

In Rogoff’s model, for example, central banks target employment and price stability, but their employment target (though socially optimal) is higher than that of wage setters. Wage setters are able to frustrate any effort by their country’s central bank to raise the level of employment by setting base nominal wages “at a sufficiently high level so that, in the absence of disturbances, the central bank will not choose to inflate the money supply beyond the point consistent with wage setters’ desired real wage” (Rogoff (1985a, p. 202)). When the two countries’ central banks act noncooperatively, they are constrained by the fact that if one inflates, the positive effect on employment will be mitigated by a depreciation of the exchange rate, which, given indexation of wages to the consumer price index, will mainly show up in higher inflation. If both countries increase the money supply, however, the exchange rate will not change, and the stimulative employment effects will be larger. If they can agree to cooperate, then, the outcome will be a higher global inflation rate, but a better response to negative output disturbances in each country. Because these two effects go in opposite directions, it is quite possible that welfare in one or both countries may be lower than in the noncooperative solution.

A more extreme view of the negative aspects of coordination denies that government policy choices are based on attempts to maximize a country’s welfare. In this view, policymakers have their own objective of personal power and profit, as do the bureaucrats that negotiate and implement agreements; policy coordination is viewed as a process of collusion (Vaubel (1985, p. 235)). In this view, cooperation may enhance the chances of politicians to be re-elected, because of publicity given to Economic Summits and the possibility of shirking some of the blame for unpopular decisions, but it does nothing to increase welfare in the countries concerned. Although this cynical view of government behavior is a good antidote to the assumption that social welfare is necessarily of paramount importance, it seems too negative an assessment. In any case, its main import concerns possible reform of the electoral process and the public choice problem (Brennan and Buchanan (1980)), rather than the issue of international cooperation.

Empirical Evidence

Even if there is a strong theoretical case for expecting policy coordination to lead to a better global economic outcome (at least when governments try to maximize the welfare of their citizens and the effects of policies are well understood), the magnitude of those possible gains is an empirical question. To date there have been relatively few empirical studies, to a large extent because technical issues concerning methods of solution and time inconsistency had to be faced first.

The conclusion that emerges from the empirical studies that have been published is clearly that gains from coordination are small (Oudiz and Sachs (1984) and Taylor (1985)).20 However, the assumptions made in these studies that information is freely available and that there is a consensus on how the global economy operates are serious limitations. Canzoneri and Minford (1988, forthcoming) have argued that the real issue is whether governments take account of the best available information on foreign policy actions in making their own choices, not whether they formulate policy jointly. The Nash equilibrium assumes that a country takes into account policies chosen abroad (but assumes that they do not respond to the home country’s choices). Canzoneri and Minford contrasted this with the “insular” solution, in which governments take no account of what other governments are doing, in the sense that they assume that other governments’ actions reflect their average behavior in the past. They argued that such a model better describes the actions of France and other European governments in the first part of the 1980s. They showed that in the Liverpool World Model (developed at the University of Liverpool, United Kingdom) insular behavior has a strongly negative effect on the welfare of industrial countries.

The issue of lack of consensus about the correct model of the economy was considered by Frankel and Rockett (1986). They supposed that eight international macroeconomic models represented at a Brookings Institution conference that took place in March 1986 (“Empirical Macroeconomics for Interdependent Economies”; see Bryant and others (1988)) are candidates for the “true” model. Coordination between the U.S. and non-U.S. central banks improves U.S. welfare in only 289 out of the 512 possible combinations of models (each of the two regions believes that one of the models is true, but the actual true model may differ from both), whereas the non-U.S. region’s welfare improves in only 297 cases. Such a comparison may give too pessimistic an assessment. Frankel and Rockett assumed that policymakers are myopic, since they all believe that they have the “truth” about the way economies work, even though they clearly do not. Ghosh and Masson (1988). however, showed that if model uncertainty is taken into account, policy will be set in a more cautious fashion, and gains from policy coordination are more likely to result (see also Holtham and Hughes Hallett (1987)). Both Canzoneri and Minford (1988, forthcoming) and Frankel and Rockett (1986) have illustrated the importance of what we have called the most basic form of cooperation, the exchange of information. Empirical knowledge of transmission mechanisms and understanding of how each country’s actions affect others are at a rudimentary stage. It is important that the implications of policy choices be understood; this is a precondition for gains to be made by jointly deciding on policies.

A General Assessment

The technical literature that we have surveyed provides a valuable insight into policymaking in interdependent economies: under some circumstances, choices made independently by national governments may not be as effective in achieving their objectives as policies that are coordinated with other governments. The practical implications of this theoretical possibility are still unclear, however, and to date they have not been fully explored in the literature. Much of the work has assumed that the structure of economies is known, and the models used have not adequately reflected the complexity of the economic system. In particular, the dangers of fine-tuning policy instruments when the precise effects of those policy instruments—and especially the lags with which they operate—are not known have not been considered. Therefore, empirical estimates of the gains to coordination are not really relevant to real-world policy choices. Confidence that coordination is likely to improve welfare requires knowledge of the functioning of the economic system. It seems that international cooperation that includes information interchange and multilateral surveillance would for the foreseeable future yield greater benefits than coordination of economic policies.

How coordination is actually to be achieved is also usually not made clear. Because there are costs to negotiating and policing agreements, it must be shown that these costs do not exceed the benefits in terms of better achievement of policy objectives. Further work is warranted on possible institutional frameworks for achieving coordination.

In many circumstances, establishing appropriate rules for the functioning of the international monetary system and the achievement of coordination of economic policies would be seen as substitutes. Given the difficulties of achieving coordination, the design of the international monetary system should have the goal of making countries choose policies on their own that are optimal from a global standpoint—or at least as close to optimal as possible (Hamada (1979)). It is unlikely to be the case, however, that any system can be designed that automatically yields the best policies in all circumstances. The present exchange rate system has stood up well to a variety of shocks since the advent of generalized floating in 1973, and it has the advantage of being flexible and adaptable. The focus of this paper is therefore mainly on how the present system can be made to function better. Specific proposals are discussed in the next section.

IV. Increasing the Effectiveness of International Cooperation

The paper has distinguished international cooperation, which includes information interchange and international surveillance, from coordination (also included in cooperation), which refers to agreements between countries to adjust their policies in light of shared objectives or to implement policies jointly. It has been argued that measures to improve international cooperation are clearly desirable, even though the case for coordination is less clear. Indeed, the consequences of coordination implemented within a wrong or ill-informed assessment of underlying economic conditions may conceivably lower the welfare of the participating countries, as discussed above. A strong case therefore exists for improving the overall information base.

Of particular interest is a recent initiative within international forums to collect quantitative “indicators” of policy actions and economic performance for major countries. The function of such indicators would be to give timely warning of the emergence of serious imbalances and to monitor progress in eliminating such imbalances. Careful analysis of these indicators might also help to suggest periods when policy coordination was desirable. A more ambitious proposal would use such indicators to trigger a change in policy if they departed from their “target zones.”

Economic Indicators

Economic indicators refer to various measures of a country’s policy actions and economic performance over a specified period. Such indicators have been in use for some time by individual countries as part of their domestic surveillance effort to assess and monitor their economic performance. Their use in policy coordination derives from the request by the Interim Committee in April 1986 that a study be undertaken of the possible use of indicators in a medium-term framework, and for the purposes of multilateral surveillance (Crockett and Goldstein (1987, pp. 34–43)). In the communiqué issued at the close of the April 1988 meeting of the Interim Committee. “Committee members stressed the importance of policy coordination in strengthening economic performance. In this context, they welcomed the progress that had been made, within the framework of the Fund’s responsibilities for surveillance over members' economic policies, in developing the use of economic indicators in a medium-term context” (IMF Survey, Vol. 17 (April 18, 1988, p. 114)).

Within the context of the present discussion, it is worth considering more specifically how the further development of a set of economic indicators might increase the effectiveness of international cooperation. Section I identified three sources of the need for coordination: (1) ignorance on the part of policymakers of the intentions of the authorities in other countries and of the implications of their decisions, (2) externalities created by trade and financial links between countries, and (3) global inconsistencies related to the n th-country problem. In addition to their role as a guide to domestic policy action, the development and analysis of a set of economic indicators might enable a clearer and more timely identification of the need for international coordination emanating from any of these sources. This in turn supports the case for the inclusion of a broad range of indicators, encompassing both policy instruments and targets, A broad range of indicators would supplement the uncertain link between exchange rate movements and the setting of monetary and fiscal policies, thereby reducing the likelihood of misleading signals. Further, if indicators are to act as a warning signal to the authorities, sole reliance upon data with long time lags, such as current account imbalances and output, is likely to be insufficient.

The collection of data alone would be inadequate for the purposes discussed above unless it were also accompanied by an analytical framework within which short- and medium-term developments can be assessed. One suggested application of economic indicators is to compare the international inconsistencies over the medium term of national projections. Early warning signals could lead to actions to avoid a possible conflict in policy objectives. Another use would be to focus attention on the possible unsustainability of current account developments over the medium term. The specific use of indicators to highlight global inconsistencies could fulfill a further function: an appropriate set of indicators could serve as a signaling device to the authorities to identify both those unsustainable policies and developments in individual countries that might be corrected by policies pursued independently and those that would benefit from some form of policy coordination.

In summary, the further development of economic indicators—including establishing different sets of indicators to serve different purposes—could fill an important role in international surveillance. They could help to identify serious imbalances affecting one or several countries, requiring the independent adjustment of national policies and possibly also international coordination of those policy adjustments. By acting as an objective signal of the need for multilateral discussion, indicators would strengthen the consultation process and the information exchange among countries that is critical to effective cooperation.

Target Zones

Proposals to establish target zones focus primary (although not exclusive) attention on the role of movements in exchange rates for signaling the need for offsetting adjustment of macroeconomic policies (Frenkel and Goldstein (1986)). Such proposals differ from proposals that call for a return to a regime of fixed exchange rates because, although the authorities would intervene in the foreign exchange market to maintain exchange rates within a certain zone of fluctuation, there may be no formal commitment to do so (Williamson (1983) and Williamson and Miller (1987)). In the so-called “hard” version of target zones, monetary policy is geared toward maintaining exchange rates within a narrow, infrequently revised, and publicly announced zone. In the “soft” version, more limited attention is paid by the monetary authorities to the exchange rate, and the zones are fairly wide, frequently revised, and confidential.

An exchange rate rule may act as an effective mechanism for achieving monetary discipline, as illustrated earlier by the experience of countries whose currencies participate under the ERM of the EMS. The experience of this group of countries and of the major industrial countries that operate under floating rates shows, however, that this mechanism is not necessary for reducing inflation, nor is it sufficient for eliminating current account imbalances. It can be argued that, since the exchange rate is a powerful transmission channel between countries, undesirable movements in both the level and variability of exchange rates may provide a useful measure of undisciplined policy actions. But the exchange rate is not the only channel of transmission of macroeconomic policies. Hence, fixing exchange rates will not remove all the sources of interdependencies between countries. The role of the exchange rate in the transmission mechanism will differ according to the type of disturbances to the domestic economy. For example, with some shocks greater fixity of exchange rates may intensify suboptimal outcomes that are the result of decentralized decision making, whereas for other shocks target zones may lead economies closer to a Pareto-optimal outcome (Gavin (1986)).

It can also be argued that exclusive focus on exchange rate movements with relation to a specified target zone may not be a sufficient statistic for the purposes of international surveillance. For example, current Fund practice is to use a range of domestic indicators to monitor economic policies. Use of the exchange rate as the primary indicator of disequilibria in underlying macroeconomic policies could send misleading signals, and thereby elicit an inappropriate policy response. Target zone proposals also require resolving the issue of the appropriate policy instruments that should be used in response to any departure from the zone. The difficulty here is that appropriate measures depend on identifying the source of exchange rate movements. For instance, a sustained real appreciation of the exchange rate may reflect a structural fiscal deficit or excessive real wages; in both cases a monetary response would be inappropriate. Even if the underlying causes of exchange rate movements can be identified, during periods of major disturbances and shifts in saving-investment balances the required change in target zones may be impractical to implement.

In summary, it appears that the emphasis placed on exchange rates as a trigger for appropriate macroeconomic policies may be misplaced. Nevertheless, one important merit of this proposal needs to be recognized: the desirability of paying greater attention to exchange rates in the conduct of macroeconomic policy. When combined with the use of other indicators, explicit attention to the consistency of exchange rates is likely to lessen the resort to beggar-thy-neighbor policies that might otherwise result in undesirable world outcomes.

V. Conclusions

This study has considered an array of important issues that relate to international cooperation and policy coordination with a view toward assessing the scope and limits to both that exist within the present international monetary system. In the discussion, the term policy coordination has been used mainly in the sense understood by policymakers; that is, it refers to agreements between countries to adjust their policies in the light of shared objectives or to implement policies jointly. Economic cooperation was defined more broadly to include both coordination and other activities such as consultation and information exchange among countries.

The perceived need for cooperation and coordination were identified as arising from the various interdependencies that exist among countries. One of the main implications of these interdependencies and the considerable uncertainty pertaining to developments and policies in other countries is that there are clearly benefits, of potentially large but unknown magnitude, to be derived from international cooperation in the form of information exchange and international surveillance. A few empirical studies have attempted to measure the gains from policy coordination and have concluded that the gains are relatively modest. This conclusion can be only tentative, however, because a variety of issues in the assessment remain unresolved. In the present international monetary system, it is likely that international cooperation can be enhanced by an analytical framework and information base that permits the identification of existing or potential imbalances that may seriously affect one or several countries. An example of such an initiative is work being undertaken at the International Monetary Fund on indicators of policy actions and economic performance for major countries. Progress in this area may also serve to identify the need for policy coordination, and to increase the likelihood of its being effective in improving the functioning of the international monetary system.


    ArgyVictor and JoanneSalop“Price and Output Effects of Monetary and Fiscal Policy Under Flexible Exchange Rates,”Staff PapersInternational Monetary Fund (Washington) Vol. 26 (June1979) pp. 224l56.

    ArtisMichael“The European Monetary System: An Evaluation,”Journal of Policy Modeling (New York) Vol. 9 (Spring1987) pp. 17598.

    ArtisMichael and SylviaOstryInternational Economic Policy Coordination Chatham House Papers No. 30 Royal Institute of International Affairs (London: Routledge and Kegan Paul1986).

    BoughtonJames M.Richard D.Haas and Paul R.Masson“The Role of Exchange Rate Movements in Transmitting International Disturbances,”paper presented at a conference on Domestic and International Aspects of Macroeconomic Policy, held in Urbana-ChampaignIllinoisSeptember 12–131986;inD. R.Hodgman and GeoffreyWoodeds. Macroeconomic Policy and Economic Interdependence (New York: Macmillanforthcoming1988).

    BrennanGeoffrey L. and JamesBuchananThe Power to Tax: Analytical Foundations of a Fiscal Constitution (Cambridge: Cambridge University Press1980).

    BryantRalph and others eds. Empirical Macroeconomics for interdependent Economies,2 vols. (Washington: The Brookings Institution1988).

    BuiterWillem H. and RichardMarstoneds. International Economic Policy Coordination (Cambridge, England: Cambridge University Press1985).

    CanzoneriMatthew B. and Dale W.HendersonStrategic Aspects of Macroeconomic Policymaking in Interdependent Economies (Washington: The Brookings Institutionforthcoming1988).

    CanzoneriMatthew B. and PatrickMinford“Policy Interdependence: Does Strategic Behaviour Pay? An Empirical Investigation Using the Liverpool World Model,”in D. R.Hodgmanand GeoffreyWoodeds. Macroeconomic Policy and Economic Interdependence (New York: Macmillan forthcoming 1988); originally appeared as Discussion Paper 201 (London: Centre for Economic Policy ResearchOctober1987).

    CooperRichard N. (1986a) “Economic Interdependencies and Coordination of Policies,” inRichard N.CooperEconomic Policy in an Interdependent World (Cambridge, Massachusetts: MIT Press) pp. 289331;originally published inR. W.Jones and P. B.Keneneds. Handbook of International EconomicsVol. 2 (Amsterdam: North-Holland; New York: Elsevier1985).

    CooperRichard N. (1986b) “International Cooperation in Public Health as a Prologue to Macroeconomic Cooperation,”Brookings Discussion Papers in International Economics No. 44 (Washington: The Brookings InstitutionMarch).

    CordenW. Max“Macro-economic Policy Co-ordination”in W. MaxCordenInflation Exchange Rates and the World Economy: Lectures on International Monetary Economics3rd ed. (Oxford: Clarendon1985) pp. 18090.

    CordenW. Max and StephenTurnovsky“Negative International Transmission of Economic Expansion,”European Economic Review (Amsterdam) Vol. 20 (January1983) pp. 289310.

    CrockettAndrew and MorrisGoldsteinStrengthening the International Monetary System: Exchange Rates Surveillance and Objective Indicators Occasional Paper 50 (Washington: International Monetary FundFebruary1987).

    CurrieDavidPaulLevine and NieVidalis“International Cooperation and Reputation in an Empirical Two-Bloc Model” in Global Macroeconomics Policy Conflict and Cooperationed. byRalphBrvant and RichardPortes (New York: St. Martin’s1987) pp. 75127.

    DornbuschRudiger“Expectations and Exchange Rate Dynamics,”Journal of Political Economy (Chicago) Vol. 84 (December1976) pp. 116176.

    EichengreenBarry“International Policy Coordination in Historical Perspective: A View from the Interwar Years” in International Economic Policy Coordinationed. byWillem H.Buiterand Richard C.Marston (Cambridge: Cambridge University Press1985) pp. 13981.

    FlemingJ. Marcus“Domestic Financial Policies Under Fixed and Floating Exchange Rates,”Staff PapersInternational Monetary Fund (Washington) Vol. 9 (November1962) pp. 36980.

    FrankelJeffrey A. and KatharineRockett“International Macroeconomic Policy Coordination when Policy-Makers Disagree on the Model,”NBER Working Paper 2059 (Cambridge, Mass: National Bureau of Economic ResearchOctober1986).

    FrenkelJacob A.“International Interdependence and the Constraints on Macroeconomic Policies,”Weltwirtschaftliches Archiv (Kiel) Vol. 122 (No. 41986) pp. 61544.

    FrenkelJacob A.and MorrisGoldstein“A Guide to Target Zones,”Staff PapersInternational Monetary Fund (Washington) Vol. 33 (December1986) pp. 63373.

    GavinMichael“Macroeconomic Policy Coordination Under Alternative Exchange Rate Regimes” (unpublished; Washington: Board of Governors of the Federal Reserve SystemSeptember1986).

    GhoshAtish R. and Paul R.Masson“International Policy Coordination in a World with Model Uncertainty,”Staff PapersInternational Monetary Fund (Washington) Vol. 35 (June1988) pp. 23058.

    GiavazziFrancesco and AlbertoGiovannini“The EMS and the Dollar,”Economic Policy: A European Forum (Cambridge, England) Vol. 1 (April1986) pp. 45585.

    HamadaKoichi“Alternative Exchange Rate Systems and the Interdependence of Monetary Policies” inR.Z.AliberNational Monetary Policies and the International Financial System (Chicago: University of Chicago Press1974) pp. 1333.

    HamadaKoichi“Macroeconomic Strategy and Coordination Under Alternative Exchange Rates”in RudigerDornbuschand Jacob A.Frenkeleds. International Economic Policy: Theory and Evidence (Baltimore: The Johns Hopkins University Press1979) pp. 292324.

    HolthamGerald and AndrewHughes Hallett“International Policy Coordination and Model Uncertainty”ed. byRalphBryant and RichardPortesinGlobal Macroeconomics: Policy Conflict and Cooperation (London: Macmillan1987) pp. 12884.

    International Monetary FundArticles of Agreementamended effective April 1978; second amendment adopted April 30 1976 (Washington: International Monetary Fund1978).

    International Monetary Fund (1984a) Issues in the Assessment of the Exchange Rates of Industrial CountriesOccasional Paper 29 (WashingtonJuly).

    International Monetary Fund (1984b) The Exchange Rate System: Lessons of the Past and Options for the FutureOccasional Paper 30 (WashingtonJuly).

    International Monetary FundInternational Financial Statistics,Vol. 40 (February1987)and various issues.

    KnightMalcolm D. and JoanneSalop“The New International Monetary System: Some Issues”inThe Contemporary International Economy: A Readered. byJohnAdams (New York: St. Martin’s Press1979) pp. 24148.

    KydlandFinn and EdwardPrescott“Rules Rather than Discretion: The Inconsistency of Optimal Plans,”Journal of Political Economy (Chicago) Vol. 85 (June1977) pp. 47391.

    McKibbinWarwick“Summary of Papers and Proceedings of the Brookings Workshop on Intergovernmental Consultations and Cooperation about Macroeconomic Policies,”Brookings Discussion Papers in International Economics No. 36 (Washington: The Brookings InstitutionSeptember1985).

    McKibbinWarwick and JeffreySachs“Comparing the Global Performance of Alternative Exchange Arrangements,”Brookings Discussion Papers in International Economics No. 49 (WashingtonThe Brookings InstitutionAugust1986).

    McKinnonRonaldAn International Standard for Monetary Stabilization (Washington: Institute for International Economics1984).

    MelitzJacques“The Welfare Case for the European Monetary System,”Journal of international Money and Finance (Guildford England) Vol. 4 (December1985) pp. 485506.

    MundellRobert“The Appropriate Use of Monetary and Fiscal Policy for Internal and External Stability,”Staff PapersInternational Monetary Fund (Washington) Vol. 9 (1962) pp. 7079.

    MussaMichael“Macroeconomic Interdependence and the Exchange Rate Regime”in RudigerDornbuschand Jacob A.Frenkeleds. International Economic Policy: Theory and Evidence (Baltimore: The Johns Hopkins University Press1979) pp. 160204.

    OudizGilles and JeffreySachs“Macroeconomic Policy Coordination among the Industrial Economies”Brookings Papers on Economic Activity: 1 (1984) The Brookings Institution (Washington) pp. 175.

    PolakJacques J.Coordination of National Economic Policies Occasional Paper 7 (New York: Group of Thirty1981).

    PutnamRobert D. and NicholasBayneHanging Together: The Seven-Power Summits (Cambridge, Massachusetts: Harvard University Press1984).

    RogoffKenneth (1985a) “Can International Monetary Policy Cooperation be Counterproductive?”Journal of International Economics (Amsterdam) Vol. 18 (May1985) pp. 199217.

    RogoffKenneth (1985b) “Can Exchange Rate Predictability be Achieved without Monetary Convergence? Evidence from the EMS,”European Economic Review (Amsterdam) Vol. 28 (June–July1985) pp. 93115.

    SachsJeffrey“Wages Profits and Macroeconomic Adjustment: A Comparative Study”Brookings Papers on Economic Activity: 2 (1979) The Brookings Institution (Washington) pp. 269319.

    SachsJeffrey“The Dollar and the Policy Mix: 1985”Brookings Papers on Economic Activity: 1 (1985) The Brookings Institution (Washington) pp. 11797.

    ShaferJeffrey R. and Bonnie E.Loopesko“Floating Exchange Rates After Ten Years”Brookings Papers on Economic Activity: 1 (1983) The Brookings Institution (Washington) pp. 170.

    SteinherrAlfred“Convergence and Coordination of Macroeconomic Policies: Some Basic Issues”European EconomyCommission of the European Communities (Brussels) No. 20 (July1984) pp. 69110.

    TaylorJohn“International Coordination in the Design of Macroeconomic Policy Rules,”European Economic Review (Amsterdam) Vol. 28 (June–July1985) pp. 5381.

    UngererHorst and others The European Monetary System: Recent DevelopmentsOccasional Paper 48 (Washington: International Monetary FundDecember1986).

    VaubelRoland“International Collusion or Competition for Macroeconomic Policy Coordination? A Restatement,”Recherches Economiques de Louvain (Louvain Belgium) Vol. 51 (December1985) pp. 22340.

    WilliamsonJohnThe Exchange Rate SystemPolicy Analyses in International Economics No. 5 (Washington: Institute for International Economics1983).

    WilliamsonJohnand Marcus H.MillerTargets and indicators: A Blueprint for the international Coordination of Economic PolicyPolicy Analayses in International Economics No. 22 (Washington: Institute for International Economics1987).

Ms. Home, a Senior Economist in the Research Department, received her Ph.D. from the Graduate Institute of International Studies, University of Geneva.

Mr. Masson, an Advisor in the Research Department, received his Ph.D. from the London School of Economics and Political Science.

The authors are indebted to their colleagues in the Fund and to Jeffrey Frankel, Doug Purvis, and David Vines for their helpful comments.

See for example, Artis and Ostry (1986, p. 14).

The term “coordinated intervention” in exchange markets by central banks is a well-known example of this broader usage.

See Artis and Ostry (1986, p. 21), and Steinherr (1984, p. 73).

The latter term is due to Artis and Ostry (1986).

Of course, foreign repercussions may not be given much weight in policymakers’ welfare calculations, but this is a separate issue.

Compatibility need not, however, ensure optimality.

Under its present Articles of Agreement, the Fund is required to “oversee the international monetary system in order to ensure its effective operation.” See Article IV, Section 3(a and b) of the Fund’s Articles (International Monetary Fund (1978, p. 8)).

As of December 31, 1986, 91 countries operated under a pegged exchange rate arrangement, 13 countries had adopted a limited flexibility arrangement, and 46 countries had adopted a flexible exchange rate arrangement (International Monetary Fund, 1987).

The post-Bretton Woods experience has been studied by, among others, Knight and Salop (1979), Shafer and Loopesko (1983), Williamson (1983), International Monetary Food (1984b), and Sachs (1985).

For a discussion of these issues, see International Monetary Fund (1984a, b).

The concept of equilibrium payments balances takes into account, among other factors, the sustamability of positions over the medium term. For a detailed discussion of underlying balance of payments positions, see International Monetary Fund (1984a).

This table may be considered as an updated version of one in International Monetary Fund (1984b, p. 27).

The group of participating countries are Belgium-Luxembourg, Denmark, France, Federal Republic of Germany, Ireland, Italy, and the Netherlands.

The EMS may be modeled as a cooperative game that yields benefits to its members that derive both from fixing the exchange rate and from allowing realignments to occur in response to shocks (see Melitz (1985) and Giavazzi and Giovannini (1986)).

In general, coordinated policies need not be directed toward the goal of convergence, and convergence need not result from coordination. The convergence (for instance, of living standards) may require differential growth rates and hence divergence of policies over some adjustment period.

See Mussa (1979) and Boughton, Haas, and Masson (1988) for surveys of that literature.

It is also possible that a large share of the gains from Stackelberg leadership by a large country may accrue to the smaller countries (see Eichengreen (1985)).

The free-rider problem also applies to the use of appreciation to achieve a reduction of inflation discussed above. In this case, each country may run a contractionary monetary policy but want other countries to expand their money supplies, allowing the first country to achieve a larger appreciation.

A more recent study (Currie, Levine, and Vidalis (1987)) has claimed that this negative assessment results from comparison of time-consistent policies only—that is, it is assumed that governments cannot carry out fully optimal policies because of lack of credibility. The work of these authors suggests that noncoordinated policies may be significantly worse than coordinated ones when governments are able to implement time-inconsistent policies.

Other Resources Citing This Publication