In this fortieth anniversary year of the international monetary conference at Bretton Woods, New Hampshire, there have been numerous but vague calls for a new Bretton Woods conference to improve our international monetary system which, if not actually ailing, at least leaves many participants uneasy and discomfited. Much of the discomfort relates to the large and burdensome external debt that has accumulated around the world, but much also goes beyond debt to the underlying monetary arrangements among countries.

Are international monetary arrangements stable? Are they likely to survive over a considerable period of time, such as a couple of decades? My answer is negative. Dissatisfaction with the very short-run and year-to-year movements in real exchange rates, combined with technological developments which will lead to further integration of the world economy, will sooner or later force a change of existing arrangements. Unless that alteration is carefully managed, it will take the form of defensive, insulating measures involving restrictions on international transactions, both trade and finance. That would be politically divisive and economically costly.

A new Bretton Woods conference is wholly premature. But it is not premature to begin thinking about how we would like international monetary arrangements to evolve in the remainder of this century. With this in mind, I suggest a radical alternative scheme for the next century: the creation of a common currency for all of the industrial democracies, with a common monetary policy and a joint Bank of Issue to determine that monetary policy. Individual countries would be free to determine their fiscal policy actions, but those would be constrained by the need to borrow in the international capital market. Free trade is a natural but not entirely necessary complement to these macroeconomic arrangements.

This suggestion, outlined in the following pages, is far too radical for the near future. It could, however, provide a "vision" or goal that can guide interim steps in improving international monetary arrangements, and by which we can judge the evolution of national economic policy.

The plan of the article is as follows. Section II offers a brief sketch of the main features of the Bretton Woods system and why it failed, drawing attention to two intrinsic flaws in the original conception. Section III briefly characterizes the present system and suggests that it is workable and even useful, but unstable in the long run-again, it suffers from two fundamental weaknesses. Section IV offers a technically workable scheme for the twenty-first century which, however, calls for major political commitments to international collaboration by the key countries, commitments which are much too ambitious for the present time. Section V brings us back to the present and suggests what steps we might take in the near future with a view to reaching the longer-term objective as it becomes politically possible.


Bretton Woods. The system that emerged from the Bretton Woods Conference 40 years ago had five key structural features:

First, it consciously provided a great deal of freedom for national economic policy to pursue national economic objectives, with the objective of assuring full employment, price stability, economic growth, and so forth. The Bretton Woods agreement was produced in the same climate of opinion which resulted in the White Paper on Employment Policy in the United Kingdom, the Employment Act of 1946 in the United States, and comparable legislation or statements of national policy in other countries, deriving directly from the experience of the 1930s and from a determination that that experience should never be repeated.

Second, the Bretton Woods system stipulated that exchange rates between currencies should be fixed. It was taken for granted that fixed exchange rates were desirable against the turbulent background of flexible exchange rates that prevailed in the early 1920s and again briefly in the early 1930s.

Third, currencies should be convertible one into another for international trade in goods and services, including travel. Again, that stipulation was against the background of extensive use of exchange controls by Nazi Germany during the 1930s and the tight wartime restrictions on trade and payments levied by many countries, which the Bretton Woods architects considered desirable to end as quickly as possible.

These three features taken together-autonomy of national policies, fixed exchange rates, and convertibility of currencies-were in conflict with one another. Countries could not frame their national economic policies independently and still maintain fixed exchange rates and currency convertibility except by luck and coincidence. That potential conflict was recognized by the Bretton Woods architects, who therefore added two further features.

Fourth, provision was made for medium-term international lending to cover balance-of-payments deficits that might result temporarily from the combination of the first three features. A new institution, the International Monetary Fund (IMF), was created as a channel for this new lending.

Fifth, countries were allowed, and in time came to be encouraged, to alter their exchange rates if it became clear that imbalances in payments were not temporary in nature. In other words, if a "fundamental disequilibrium" emerged, the exchange rate was to be changed by a discrete amount, with international agreement, in recognition that such imbalances would be inappropriate to finance indefinitely.

These then were the basic features of the Bretton Woods system. Interestingly, there was no provision in the system for secular growth in international reserves beyond a somewhat ambiguous provision permitting what was called a "uniform change in par values": that is to say, a deliberate rise in the price of gold. It was implicitly assumed that new gold production taken into monetary reserves would be sufficient to provide for adequate growth. In the event, the U.S. dollar came to provide for the needed liquidity, as well as emerging as the currency of intervention in a regime in which some operating medium was necessary to assure that exchange rates remained fixed.

During the quarter century between 1945 and 1970, world monetary reserves outside the United States grew by $54 billion, averaging 4.5 percent per year. Gold provided $13 billion of this increase, of which $9 billion was transferred from the high gold reserves of the United States (which reached 70 percent of total world monetary gold reserves in the late 1940s), and $4 billion came from new gold production. Foreign exchange, which was overwhelmingly in dollars, provided $30 billion of the growth in reserves. The IMF provided $11 billion, including $3 billion of the new Special Drawing Rights (SDRs) in 1970. U.S. reserves, of course, declined during this period because a substantial part of its gold stock was lost to other countries.

As it emerged-though not as it was designed-the Bretton Woods system might be said to have involved a bargain between the United States and the rest of the world: the United States would maintain domestic economic stability, and other countries would fix their currencies to the dollar and would accumulate their reserves in gold-convertible dollars. After a relatively brief period of postwar redistribution of the world's monetary gold stock, they would not actually convert their dollars into gold. Under this bargain, other countries would import economic stability from the United States. If a country got out of line with the world norm, it would have to change the par value of its currency. In turn, the United States did not have to be as concerned as other countries about how it financed a balance-of-payments deficit. Indeed, the very notion of a balance-of-payments deficit was an ambiguous one for the United States under these circumstances, although that did not keep the Commerce Department from publishing figures which it called the "deficit" for many years.

A second characteristic of this arrangement was that the dollar was overvalued relative to what it would have been without steady accretion of dollars in the reserves of other countries. That feature permitted some export-led growth by the rest of the world that would not have taken place under different monetary arrangements, under which American products would have been somewhat more competitive in world markets.

In this view of the world, the United States broke its part of the bargain in the late 1960s by inflating too much in connection with the Vietnam War and the Great Society programs. Some Europeans thought that the United States was inflating too much even in the early 1960s. On this point, they would not have found much agreement from Americans. Indeed, disagreement over U.S. policy in the early 1960s indicated one of the weaknesses of the supposed bargain, namely dispute around the world over what exactly represented economically stabilizing behavior by the United States.

The structure of the Bretton Woods system had two intrinsic flaws in it, so that it would have broken down sooner or later even without the burst of U.S. inflation in the late 1960s. First, the gold convertibility of the dollar was bound to become increasingly doubtful as dollar liabilities rose over time relative to the U.S. gold stock. To halt the accumulation of foreign-held dollar reserves would have stifled growth of the world economy. Yet to allow the accumulation to continue would have gradually undermined the foundation of the system. Professor Robert Triffin of Yale University pointed out this dilemma as early as 1959. A new international reserve asset, IMF-issued Special Drawing Rights-aptly described as paper gold at the time-was finally created in the late 1960s as a long-run substitute for the dollar, thus offering a solution to the dilemma. But the solution came too late. This part of the system broke down in 1971, when President Nixon indefinitely suspended gold convertibility of the dollar. Two points are worth noting in passing. The first is that the U.S. dollar was the only currency that was convertible into gold, even though the Bretton Woods agreement was formally symmetrical with regard to all currencies. The second is that countries continued to accumulate dollars in their international reserves even after gold convertibility of the dollar was suspended.

The second flaw in the Bretton Woods system was its reliance on discrete changes in exchange rates to correct imbalances in payments. Once a disequilibrium persisted long enough to be "fundamental" rather than temporary in nature, it was clear to everyone and the system thus produced the celebrated one-way option for currency speculation. Since the remedy to a fundamental disequilibrium was a jump in the value of the currency, speculators could move into or out of the currency at relatively low cost when they thought the jump would occur and take their gains after it occurred. It is interesting to note that the Bretton Woods architects had appreciated this problem, at least in principle, and to remedy it had stipulated that currencies should be convertible for current account transactions, but not for capital account transactions. The possibility was envisioned that countries might maintain controls on capital flows, and indeed countries were even enjoined to help other countries maintain and enforce their capital controls. So capital controls were in principle allowed under the Bretton Woods system, and indeed in a certain sense they were required by the internal logic of the system.

This feature of the system did not anticipate the enormous changes both in the nature of trade and in international capital movements that would take place over time. With improved and cheaper communications, it became easy to move capital through telegraphic transfers around the world at relatively low cost. In addition, many firms, especially American firms, began to invest heavily abroad in the 1950s and 1960s, so that many intra-corporate transactions became international in nature. Finally, international trade gradually evolved away from traditional commodity trade toward special order and long-lead-time manufactures in which payments for trade and credit terms become inextricably mixed. For all of these reasons, it became increasingly difficult to separate capital from current account transactions and to maintain control on capital transactions.

The movement of funds that was associated with anticipated discrete changes in exchange rates became quite enormous and greatly complicated the management of domestic economic policies. In many countries, they threatened the autonomy of domestic national policy which was to have been preserved by the Bretton Woods system. For example, Germany in 1969 experienced a 25-percent increase in its money supply in a single week due to the inflow of speculative funds across the foreign exchanges and to the requirement that Germany maintain a fixed exchange rate for the mark in terms of other currencies.

In truth, the free movement of capital is incompatible with a system of exchange rates that are occasionally changed by consequential amounts and in a predictable direction. This part of the Bretton Woods system broke down definitively in 1973, although the breakdown started in 1970 when Canada allowed its currency to float.

The U.S. inflation of the late 1960s resulted in large dollar outflows in the early 1970s that strained the Bretton Woods system to the breaking point. But it should be clear by now that this was only the proximate cause of the breakdown of the Bretton Woods system. The intrinsic flaws in the system would have come to the surface sooner or later, in response to one strain or another. They happened to come to the fore in 1971-73.

It is worth remarking that the breakdown of the Bretton Woods system was only partial. The International Monetary Fund is an important survivor, both as a lender and as a forum for managing the international monetary system. The convertibility of currencies and the continuing autonomy of national economic policies-both features of the Bretton Woods architecture-are still taken as desiderata in a well-functioning international monetary system. It is a measure of the success of that system that we take them for granted. It was the exchange rate features of the system that broke down, and the psychologically important but technically tenuous link to the historic gold standard via the gold convertibility of the leading currency.


Present Arrangements. For the past decade, the world has permitted a variety of exchange rate arrangements, but in practice with a much higher degree of flexibility than prevailed under the Bretton Woods system. This "non-system" has served the world economy rather well during a turbulent decade. It is true that the overall economic performance during the past decade, whether measured in terms of inflation rates or growth rates or unemployment rates, has been far inferior to what it was during the 1950s and 1960s. But it probably would have been even worse if governments had tried to maintain the Bretton Woods system through the period. In view of the large disturbances which the world economy has undergone, an attempt to maintain fixed but adjustable exchange rates would almost certainly have required a much higher degree of restrictions over both capital and current transactions than in fact prevailed. Thus exchange rate flexibility helped to preserve a relatively open trading and financial system.

During the decade, movable exchange rates have generally corrected for differences in national inflation rates, as economists predicted they would. But the movements in exchange rates have gone beyond that and affected "real" exchange rates as well-that is, competitiveness as measured by the relative prices at which the goods of one country on the average trade against the goods of another.

An evaluation of the period as a whole is complicated and difficult. Many of the movements in real exchange rates followed textbook predictions, responding to imbalances in current accounts, or to dramatic changes in resource endowments (such as the discovery of North Sea oil), or they followed divergent movements in aggregate demand. But some of the movements in real exchange rates have not followed textbook patterns, and even when they have, they have often been viewed as unwelcome disturbances by some countries, especially following the sharp depreciation of the U.S. dollar in 1978, and again following the sharp appreciation of the dollar in 1981 and 1982. Perhaps for this reason, most countries of the world in fact have not allowed their currencies to float. Rather, they have fixed the value of their currencies against something-against another currency, or a basket of currencies, or, in the case of the European Monetary System, against one another. Thus it is not entirely accurate to characterize current arrangements as involving floating exchange rates. In practice, the exchange rates of several major currencies-the U.S. dollar, the Japanese yen, the British pound, the Canadian dollar-do float more or less freely, but other currencies do not float, although they have shown greater flexibility than they would have under a Bretton Woods regime.

Movements in some key bilateral exchange rates, such as the dollar-deutsche mark rate, have shown sharp short-run variations on occasion during the past decade, which were not keyed to fundamental economic developments in any obvious way. There have been occasional weeks of average daily variations in excess of three percent. Why such great variability? In some respects foreign currency holdings are like any other financial asset, whose current price reflects all the information available that may have a bearing on its future value. New information may then affect market prices (in this instance exchange rates) sharply as the "market" reappraises the future in the light of new information.

This analogy to stock prices helps to explain the abruptness of some movements in exchange rates. But it hardly helps to explain month after month of sharp variability, up and down. Much "new" information, in a longer perspective, is in fact only noise, whose bearing on the price in question can reasonably be expected to be reversed in the near if not immediate future, especially since trade can eventually be expected to respond to persistent movements in exchange rates.

Abrupt up-and-down shifts in exchange rates may not, by themselves, greatly affect trade and production, since they should reasonably be expected to be reversed soon if they are not clearly linked to more fundamental economic developments. The added uncertainty about what an exchange rate will be when a transaction is completed will, however, undoubtedly discourage trade and investment for export to some extent.

The main difficulty with flexible exchange rates is that another influence is also at work, which can transmute the influence of noisy news into larger changes in exchange rates than otherwise would take place. It is the presence of crowd or bandwagon effects in the trading community. Few know how to interpret the news. Many use a movement in the exchange rate itself as a source of information about market sentiment. To avoid being left behind, dealers jump on the bandwagon, thus pushing the exchange rate further in the direction it tended to go initially. Expectations feed on expectations.

When this process is operating, even those who suspect the exchange rate has gone too far still have an interest in holding their investments so long as the prospect for further gain outweighs the probability of reversal. Thus a secondary judgment, oriented toward market dynamics, is superimposed on the reassessment based on the new information, and may come to dominate the movement in exchange rates for a time. This would not be troublesome if there were no real economic consequences. But in some periods expectations about the "fundamentals" may be so weakly held that the rate can be dominated by purely market dynamics for periods measured in weeks or months. When that is so, the exchange rate may in turn affect new information, such as price indexes, increases in which the public interprets as "inflation." Or it may set in motion urgent steps to avoid risks, as when multinational firms move to protect their quarterly balance sheets (at the expense of the operating earnings of the firms). So a vicious circle may temporarily be set in motion. And this vicious circle may aggravate inflation rates and hence inflationary expectations, or it may divert management attention away from real long-term investment to short-term balance-sheet considerations. In either case an unnecessary and avoidable element of instability is introduced into national economies.

Two features of present exchange rate arrangements will not be satisfactory over the long run. First, movements in real exchange rates have major effects on national economies, effects which are often unwelcome. Yet movements in real exchange rates cannot be easily controlled by use of the usual instruments of national economic policy because the determinants of exchange rates are diverse and complex. The result is that at any moment the influence of policy actions on exchange rates is uncertain. Portfolio decisions with respect to financial assets play a key role in the short-run determination of exchange rates, yet the influence of policy actions on portfolio decisions, via expectations, is uncertain. This marks a substantial contrast with the influence of policy actions on the aggregate demand for goods and services, where the linkages with policy are clearer. Despite this, we have not to date been able to eliminate the so-called business cycle. Unpredictable movements in real exchange rates, and unpredictable responses of real exchange rates to government action, greatly aggravate the problem of macroeconomic management.

Second, under a regime of flexible exchange rates there is a temptation, hence some tendency, to manipulate the exchange rate for national purposes. This can be done either to fight inflation, since monetary tightening produces an immediate reward-at the expense of other countries, so long as others do not respond in kind-in terms of a reduced inflation rate brought about by an appreciated currency. Or it can be used to combat unemployment, when expansionary monetary policy depreciates the currency-again, in general, at the expense of other countries. Of course, the new configuration of exchange rates may be satisfactory to all or most countries, but that would be a coincidence. Ordinarily these represent self-centered national actions which simply pass the problem either of inflation or of unemployment to other countries. Members of the IMF have a general responsibility to avoid such manipulation of exchange rates, and the IMF has a general responsibility for surveillance over exchange rate practices, with the aim of preventing such practices. But surveillance really has not gotten off the ground, and it is not clear under today's arrangements what the IMF can really do, for example, when a Sweden deliberately depreciates its currency in order to increase output and employment, or when a United States achieves a substantial reduction in its inflation rate through a policy of tight money which has greatly appreciated the dollar against other currencies.

Just as present exchange rate arrangements are not really sustainable over the long run, neither are present arrangements for the creation of reserves. The principal reserve medium today is a national currency, the U.S. dollar, dependent in large part for its supply on the policies of the United States. This has been accepted, more or less grudgingly, because it has worked reasonably well and there is no clear feasible alternative. But it leaves a deep sense of uneasiness around the world, even when the United States in the judgment of others is relatively well behaved; and the uneasiness grows dramatically when in such periods as 1970-71 and 1978 and 1981-82 the rest of the world, or parts of it, believe the United States is not well behaved.

Moreover, as the United States shrinks in relation to the rest of the world, as it is bound to do, the intrinsic weaknesses of reliance on the U.S. dollar will become more apparent, especially in the United States, where the possible reaction of foreign dollar-holders will become an ever greater constraint on U.S. monetary policy. The United States is bound to shrink relative to the rest of the world, not because it is doing badly, but because the rest of the world may be expected to do well. The natural growth in the labor force and the rate of capital accumulation are both higher in many parts of the world than they are in the United States. Furthermore, technologically lagging countries can reduce the technological gap between themselves and the United States, which operates on the frontiers of modern technology. Thus the simple arithmetic of economic growth will insure a gradual relative decline of the United States, for instance from about one-fourth of gross world product at present to around one-sixth 25 years from now if the United States grows on average at three percent a year and the rest of the world grows on average at five percent a year, both plausible numbers.

In short, the present set of monetary arrangements, while not in any immediate danger of collapse from its intrinsic features, as distinguished from some external event, is not stable in the long run. It is not a durable system. It must evolve into something else.

But what will or should it evolve into? One possibility is that the frustrations arising from the sense of loss of national control will lead to significant attempts to reassert national control by sharply reducing the openness and permeability of national economies to external influences. The move to flexible exchange rates can itself be interpreted as such a response, since countries enjoyed even less control, especially as regards monetary policy, under a system of fixed exchange rates with high capital mobility. But we have now learned that flexible exchange rates, while they offer some degree of greater national autonomy, do not effectively insulate national economies from external influences, and indeed in some instances may even exacerbate the impact of external influences on national economic developments. So the frustrations at loss of national control continue, and alleviating them requires much stronger insulating material than flexible exchange rates alone provide. It would probably involve a reversion to extensive use of restrictions over capital movements. And since capital transactions cannot be effectively separated from current transactions, there would be a tendency to extend restrictions to current transactions as well.


Future arrangements. I suggest a different possible evolution of international monetary arrangements that attempts to deal with the intrinsic problems with present arrangements that render them unstable in the long run. First, let us go forward 25 years, to the year 2010. That is far enough ahead so that developments that are completely unrealistic in the next five or ten years can be contemplated. But it is not so far ahead that we cannot really contemplate them at all.

By 2010 the populations and labor forces of the modern industrial economies will of course be larger than they are today, but the labor force engaged in manufacturing production in today's OECD (Organization for Economic Cooperation and Development) countries will probably have declined. Manufacturing is likely to go the way that agriculture has already gone, with a declining share of the labor force able to produce all of the material goods that the rest of society needs. Real incomes per capita will be over 50 percent higher than they are today. The world will be very electronic. Thus not only will large-scale financial transactions take place virtually instantaneously to any part of the world-we are close to that situation today-but even retail transactions in financial services and in goods can take place electronically. That is, householders will be able to purchase information regarding taxation, investments, retirement possibilities, or education by consulting electronic catalogues and information sources in their own homes. Even goods will be able to be purchased by inspecting them on a television screen, placing the order electronically, and having them delivered in a relatively short period of time. English will become even more widespread as the language of commerce.

With higher real incomes and lower relative prices for long-distance transportation, much more travel will take place than occurs today. Reliable, high-speed and low-cost communications over the globe will permit management control of production locations in many places. Lower transportation costs (relative to the price of other goods and services) will encourage trade. These factors taken together are likely to result in greater possibilities for substitution of geographic locations, not only in manufacturing production but also in many services. Thus, real movements in exchange rates will be highly disruptive of profits, production and employment in any particular location.

Yet financial factors will still dominate the determination of exchange rates in the short run. In view of the greater sensitivity of production to changes in real exchange rates, governments must reduce arbitrary movements in real exchange rates in order to maintain an open trading system. With widespread information and low transaction costs, an adjustable peg system of exchange rates that requires occasional discretionary movements in market exchange rates is not likely to be tenable-indeed, did not prove to be tenable even under the technological conditions prevailing in the 1960s.

These various considerations lead me to conclude that we will need a system of credibly fixed exchange rates by that time if we are to preserve an open trading and financial system. Exchange rates can be most credibly fixed if they are eliminated altogether, that is, if international transactions take place with a single currency. But a single currency is possible only if there is in effect a single monetary policy, and a single authority issuing the currency and directing the monetary policy. How can independent states accomplish that? They need to turn over the determination of monetary policy to a supranational body, but one which is responsible collectively to the governments of the independent states. There is some precedent for parts of this type of arrangement in the origins of the U.S. Federal Reserve System, which blended quite separate regions of the country, and banks subject to diverse state banking jurisdictions, into a single system, paralleling the increasingly national financial market. Similarly, we will need a world monetary system that parallels the increasingly global financial market. It will probably not be possible, even within the time scale envisaged here, to have a truly global Bank of Issue. But that will not be necessary either, and it may be possible to have a Bank of Issue which serves a more limited group of democratic countries, and which can serve as the core of an international system.

The Monetary Authority. The tasks, the instruments, and the decision-making structure of the Bank of Issue could look something like the following:

-The governing board would be made up of representatives of national governments, whose votes would be weighted according to the share of the national GNP in the total gross product of the community of participating nations. This weighting could be altered at five-year intervals to allow for differences in growth rates.

-The task of the monetary authority would be to stabilize the macroeconomic environment and to avoid or mitigate liquidity crises by acting as a lender of last resort, just as national central banks do today. The debate on the relative weights to be attached to output and employment as opposed to price stabilization, and on how monetary policy should actually be managed, could continue just as it does at present, without prejudice.

-The Bank of Issue would accomplish its tasks by engaging in open market operations in which it issued the new currency for the securities of member countries. It could also engage in rediscount operations, whereby it extended claims against itself in exchange for acceptable paper at the initiative of banks within the system, subject to its own acquiescence in those initiatives.

The Bank of Issue need not engage in detailed regulation of the banks throughout the system covered by the new currency. That could be left in the hands of national regulators. It would, however, probably want to issue guidelines-minimum standards-to be followed by national regulators, and to maintain enough surveillance over banks to be sure of itself when it was called upon to act as a lender of last resort.

In the first instance, open market operations by the Bank of Issue could be distributed among the securities of national governments in proportion to their voting weight (i.e., their GNP share), but over time this limitation would probably cease to be necessary as financial markets evolved and securities issued by many national governments became virtually perfect substitutes one for another. In any case, the Bank of Issue's holdings of national government securities could be altered from GNP shares via the rediscounting facility, as needed.

Seigniorage in this system would automatically be distributed to national governments as their securities were purchased by the Bank of Issue, thereby giving them the purchasing power to buy goods and services. In addition, the Bank of Issue would run profits from its interest earnings, and those could be distributed from time to time to national governments on the basis of their voting shares.

The currency of the Bank of Issue could be practically anything. Most natural would be an evolution from the present U.S. dollar, making use of the extensive dollar-based worldwide markets. But if that were not politically acceptable, it could be a synthetic unit that the public would have to get used to, just as it had to get used to the metric system when that replaced numerous national systems. The key point is that monetary control-the issuance of currency and of reserve credit-would be in the hands of the new Bank of Issue, not in the hands of any national government, no matter what the historical origin of the new currency happened to be.

National Economic Policy. The publics of the industrial democracies have placed high expectations on their national governments for economic management. Here governments are being asked to pass monetary policy to a supranational agency, the actions of which they can influence but not determine, taken one by one. Would national governments be giving up all of their macroeconomic control? The answer to this question is negative, since they could still pursue fiscal policy at the national level. What they would be giving up is monetary financing of budget deficits beyond their prorated allocation from jointly agreed open market operations. In particular, they could not engage in inflationary finance to reduce the real value of outstanding currency and debt at the national level, although the requisite majority could do so at the international level. To finance budget deficits, therefore, it would be necessary to go to the capital market. But under the envisaged circumstances there would no doubt be a very high degree of capital mobility among participants, since all securities would be denominated in a single, widely used currency. Of course, the influence of national fiscal actions on national aggregate demand would be limited by leakages abroad through demand for imports, and at the outer limits by the extent to which individual governments could borrow in the capital market.

Governments could also use their fiscal powers to attract internationally mobile firms by means of tax holidays or other fiscal incentives. These practices have already emerged as a new form of fiscal action both within countries (e.g., industrial development bonds issued by individual states within the United States) and between countries. With internationally mobile capital, these practices may indeed succeed in generating local employment in "depressed" areas without necessarily resulting in a misallocation of resources, as the burden of taxation is shifted to relatively immobile residents. Nonetheless, if these practices became too competitive among nations, they might want to put some collectively agreed limits on them, and even allow special differentiation under some circumstances, e.g., when unemployment rates were higher than some agreed norm.

One old-fashioned policy instrument for encouraging local investment and employment is the use of tariffs to discriminate against foreign goods. It would be logical if this single currency regime were accompanied by free trade, just as the dollar area within the United States is accompanied by free trade. That would also be consistent with the collaborative political spirit that would be required for the single currency regime to be established. Free trade would insure one market in goods as well as in financial instruments. The scheme would be quite workable also with modest tariffs, at or below the levels that now generally prevail among industrialized countries. But the exact nature of the commercial regime is beyond the scope of this article.

How the Regime Would Work. Governments could determine the balance between their expenditures and taxes as they do now, but beyond their pro-rated share of the Bank of Issue's open market purchases and profits they would have to borrow on the capital market to cover any budget deficits. Market access would be determined by a market assessment of the probability of repayment, which would assuredly be high within a plausible range of budgetary behavior. Both receipts and expenditures would be made in the common currency, as would the borrowing. Each country could set its own course independently, with no need for formal coordination of fiscal policy. Financial markets would "coordinate" to some extent, via interest rates, since if all governments decided to borrow heavily at once, in a period in which private demands for credit were also high, interest rates would rise and that would induce greater caution in borrowing. But the larger countries would certainly find it useful to exchange information on intentions with respect to future actions, so that each of them could take the prospective actions of others into account. This exchange would no doubt evolve over time into an iterative process that was hardly distinguishable from coordination, although in the end each country would be free to act as it saw fit.

Monetary policy would be set for the community as a whole by a board of governors, who in practice would probably be finance ministers. No single country would be in control. A weighted majority of the governors would decide the principles both to govern monetary policy (e.g., how much weight to give to monetary magnitudes as opposed to other economic variables in framing monetary policy) and with respect to actual operations. The governors in turn would be accountable to legislatures. The Bank of Issue would have a certain autonomy by virtue of not being beholden to any single legislative or executive authority. Thus it could not be manipulated for particular electoral reasons. On the other hand, its actions would be determined by a majority of officials who would be individually accountable to legislatures or executives, so that if a (weighted) majority of them desired a shift in policy, it would occur.

Balance-of-payments adjustment within this regime would be as easy, or as difficult, as it is between regions of the United States or any other large country today. The adjustment would be automatic, except insofar as it was cushioned by capital inflows induced by fiscal actions. Automatic balance-of-payments adjustment sometimes leads to unemployment, as following a shift in demand away from the products of a particular region or country. Fiscal policy in its various forms could be used to cushion such unemployment. In addition, my guess is that there would be considerable net immigration into the present industrial democracies by early in the next century, and the distribution of that flow of migrants would provide considerable flexibility to the labor force in the region as a whole.

This one-currency regime is much too radical to envisage in the near future. But it is not too radical to envisage 25 years from now, and indeed some such scheme, or its functional equivalent, will be necessary to avoid retrogression into greater reliance on barriers to international trade and financial transactions. Moreover, it is useful to have a vision to provide guidance for the steps that may be feasible in the near future. Some idea of where we would like to get to provides a sense of direction for the next steps.


Next Steps. The idea of a single currency is so far from being politically feasible at present-in its call for a pooling of monetary sovereignty-that it will require many years of consideration before people become accustomed to the idea. But the economic effect can be gradually approximated by giving greater weight to exchange rates in framing national monetary policy. Many countries-all those with fixed or semi-fixed rates-of course already do this. This injunction therefore applies mainly to the United States, Canada, Japan, the United Kingdom, and the members of the European Monetary System taken as a group. If monetary policy were governed in such a way as to limit wide swings in key exchange rates, this would tend also to reduce fluctuations in real exchange rates. This result could be accomplished by adoption of one or another of the formal schemes that have been proposed from time to time, such as the target zone, whereby countries undertake to confine market movements of the exchange rate within a specified band centered on a target rate, which target can if necessary be altered from time to time. The European Monetary System is a variant of such a scheme, with central rates being subject to periodic renegotiation as they become questionable. Seven changes in central rates have been made in the period since 1979, and generally the changes have been sufficiently small so that market exchange rates were not immediately affected, or were affected only modestly.

It may not be possible to reach international agreement on a formal scheme for exchange rate management. But the process of official discussion of such schemes, each particular one of which is subject to defects under some circumstances, will apprise officials of the possibilities for accomplishing the principal objective, viz. to reduce undue fluctuations in real exchange rates. Thus launching a move toward "reform" of exchange rate arrangements may fail in the sense that no formal scheme is agreed on, but still succeed in its underlying purpose of establishing a more or less shared view of what exchange rates should be at a given time and a consensus to work toward keeping market rates within the neighborhood of the consensus rates.

What is also necessary is some consultation among major countries on the overall "tone" of monetary policy. This is a politically difficult step and cannot be taken overtly any time soon, since each nation has its formal system of decision-making and channels of responsibilities for determining monetary policy. However, the same result can be accomplished informally, centered around discussion of exchange rate management, for which there seems to be a widespread desire, especially in business circles.

It was suggested in the previous section that the choice of a currency for a one-currency regime is open and in a sense is arbitrary. It could be anything that is agreed, since money is above all a social convention. In fact the choice would be a politically charged issue, with strong if irrational objections to the choice of any national currency. If national currencies are ruled out, that leaves the European currency unit (ECU) and SDR in today's world. The ECU might meet the same objections in the United States and Japan as the U.S. dollar would meet in Europe. That in turn leaves only the SDR, which is now defined as a weighted average of five leading national currencies: the U.S. dollar, the Japanese yen, the German mark, the French franc and the British pound. The new Bank of Issue could not issue IMF-SDRs unless the Bank were the IMF itself. But the Bank could use the SDR as its measuring unit, and issue both currency notes and reserve bank credit in that unit.

The future of the SDR as a currency would be immeasurably enhanced if private parties could conduct transactions in SDRs; indeed, that would be a necessary condition. It would also greatly facilitate the use of the SDR as a central bank currency, since the modus operandi of central banks in most cases is through private markets, and they need a medium which can be used in private markets. Thus the IMF-SDR would be enhanced if some mechanism could be found to make this possible. The IMF Articles would have to be amended to make it possible for private parties and commercial banks to hold the IMF-SDR directly. Professor Peter Kenen of Princeton University has recently made an ingenious proposal which would accomplish much the same result without formally amending the Articles.1 This is not an urgent step, but it should be done if the role of SDR is to be strengthened. Also, it would be desirable to issue more IMF-SDRs to keep that asset alive and in use. We will want it sometime in the future. The British economist John Williamson, of the Institute for International Economics, has recently shown that an issue of SDR 43 billion over the next two years, while at first glance a large figure, could easily be justified.2

A key question concerning the new Bank of Issue is what countries should participate in its management, use its currency, and forswear national monetary policy. We have come to think of the international monetary system, centered on the IMF with its 146 members, as a global system, albeit excluding most communist countries and Switzerland. That was certainly the conception at Bretton Woods, even though most of the early negotiation had been between the Americans and the British. That was also the spirit of the times at Bretton Woods, when the wartime allies placed their hopes for a better world in the United Nations Organization and its functional affiliates.

But there is serious question about whether one world money is either necessary or desirable. And it is certainly not feasible, even within our generous 25-year time frame. It is not feasible for two reasons. First, it is highly doubtful whether the American public, to take just one example, could ever accept that countries with oppressive autocratic regimes should vote on the monetary policy that would affect monetary conditions in the United States. I believe that the same reservations would obtain in other democratic societies. For such a bold step to work at all, it presupposes a certain convergence of political values as reflected in the nature of political decision-making, and the basic trust and confidence to which those give rise.

Second, countries with different values, circumstances, and systems of governance are bound to introduce into negotiations leading toward a common Bank of Issue elements which are of greater interest to them, thus broadening the agenda for negotiation and rendering impossible an already difficult negotiation. For both reasons the proposal should be undertaken in the first instance by the United States, Japan, and the members of the European Community. This group represents the core of the monetary system at present and for some time to come. Other democracies would be free to join if they wished, and if they were willing to undertake the commitments involved, but no one should be obliged to join. Very likely many countries would find it attractive in the early stages not to join, but nonetheless to peg their currencies to the SDR or whatever was the unit of account of the Bank of Issue. They would retain some monetary freedom, however, which members had given up. Some countries would also be reluctant to give up the seigniorage from currency issue, which can be consequential where currency still bears a high ratio to GNP.

In short, there would be an inner club accepting higher responsibilities, but open to additional members who met the requirements, and of value even to nonmembers by providing a stable monetary environment against which to frame their economic policies. But this arrangement would mark a formal break with the universalism that governs the de jure if not the de facto structure of the Bretton Woods system today.


1 See Peter B. Kenen, "Use of the SDR to Supplement or Substitute for Other Means of Finance," in George M. von Furstenberg, ed., International Money and Credit: The Policy Roles, Washington: IMF, 1983.

2 See John Williamson, A New SDR Allocation?, Washington: Institute for International Economics, March 1984.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Richard N. Cooper is the Maurits C. Boas Professor of International Economics at Harvard University. He was Under Secretary of State for Economic Affairs during 1977-1981, and Provost of Yale University in 1972-74. He is the author of The Economics of Interdependence and other works. This article draws on a paper prepared for the Federal Reserve Bank of Boston to honor the fortieth anniversary of the Bretton Woods Conference. Copyright (c) 1984 Richard N. Cooper.
  • More By Richard N. Cooper