1
lw du
pi T to
a
m
;it
tu a\ A
U]
bi ol
P'
Uf
aj
tc it
il si J
Ö!
r< a: tj n ti
I
il
A volume in the series
Cornell Studies in Money
edited by Eric Helleiner and Jonathan Kirshner
The Limits of Transparency: Ambiguity and the History of International Finance by Jacqueline Best
Smoke & Mirrors, Inc.: Accounting for Capitalism
by Nicolas Veron, Matthieu Autret, and Alfred Galichon translated by George Holoch
The Globalizers: The IMF, the World Bank, and Their Borrowers by Ngairc Woods
International Monetary Power
edited by
DAVID M. ANDREWS
Cornell University Press   Ithaca and London
CHAPTER TWO
The Macrofoundatlons of Monetary Power
Benjamin J. Cohen
What are the foundations of monetary power? David Andrews (chap. 1 in this volume) distinguishes between two pathways for the exercise of monetary power: the macro-level, linked to the problem of baiance-of-payments disequilibrium; and the micro-level, working through the capacity of money to alter actor interests and identities. The purpose of this chapter is to promote a clearer understanding of the sources of power at the macro-level pathway—what we may call the macrofoundations of monetary power. Building in good part on earlier contributions of my own,11 argue that the central issue at the macro-level is the distribution of the burden of adjustment to external imbalance. The macro-level dimension "of monetary power consists, first and foremost, of a capacity to avoid payments adjustment costs, either by delaying adjustment or by deflecting the burden of adjustment on to others.Ceteris paribus, the greater is a state's capacity to avoid adjustment costs, relative to that of other states, the greater is its power at the macro-level.
The devil, of course, is in the details. What do we mean by adjustments costs? What are the sources of the capacity to avoid adjustment costs—the macrofoundations of monetary power? And what are the limits of that capacity? The first of these questions is addressed in first three sections of this chapter, and the subsequent questions are addressed in sections four and five. Section six concludes.
The Burden of Adjustment
Analysis at the macro-level, I submit, must begin by focusing on the distribution of the burden of adjustment to external imbalance. The underlying source of power at
1. Including, especially, Benjamin J. Cohen, "Adjustment Costs and the Distribution of New Reserves," Princeton Studies in International Finance, no, 18 (1%6),
31
32      Benjamin J. Cohen
Macrofoitndations of Monetary Power 33
this level is a state's relative capacity to avoid adjustment costs, either by delaying the adjustment process or by deflecting the burden of adjustment to others.
Autonomy and Influence
At the most general level, power in international relations is defined as the ability to control, or at least influence, the outcome of events. In operational terms, this naturally equates with a capacity to control the behavior of actors—"letting others have your way," as diplomacy has jokingly been defined. A state, in this sense, is powerful to the extent that it can effectively pressure or coerce outsiders, in short, to the extent that it can exercise leverage or enforce compliance. As Andrews points out (chap. 1 in this volume), a common synonym for this meaning of power is, simply, influence.'1
But influence is not the only relevant meaning of power. There is also a vital second meaning, corresponding to the dictionary definition of power as a capacity for action. A state is also powerful to the extent that it is able to exercise policy inde pendence—to act freely, insulated from outside pressure in policy formulation and implementation. In this sense, power does not mean influencing others; rather, it means not allowing others to influence you—others letting you have jour way. A useful synonym for this meaning of power is autonomy.
The distinction between the two meanings is critical. Influence and autonomy may be understood as two distinct dimensions of power, which we may label, respectively, the external dimension and internal dimension. Both are based in social relationships and can be observed in behavioral terms. Both are also unavoidably interrelated. They are not, however, of equal importance. Logically, power begins with autonomy, the internal dimension. Influence, the external dimension, is best thought of as functionally derivative—inconceivable in practical terms without first attaining and sustaining a relatively high degree of policy independence at home. As the saying goes in American football, the best offense starts with a good defense. It is possible to think of autonomy without influence; it is impossible to think of influence without at least some degree of autonomy.
This does not mean that autonomy must be enjoyed in all aspects of international affairs or in all geographic relationships in order to be able to exercise influence in any aspect or relationship. Neither domain nor scope needs to be universal for power to be effective. States can successfully apply leverage in selected issue areas or rela tionships even while themselves being subject to pressure or coercion in others. But
2. The careful reader will note that, in a previous essay, I proposed the term authority rather than in flutnee for this meaning of power. See Benjamin J. Cohen, "Money and Power in World Politics," hi Strange Power; Shaping the Parameters of International Relations and International Poliikal Economy, ed. Thomas C. Lawton, James N. Rosenau, and Amy C. Verdun, 91-113 (Aldershot: Ashgftte, 2000). I am now persuaded, however, that, because of the inferences of legitimacy associated with the notion of au thority, the term influence is preferable. For more on the ties between monetary authority and legitimacy, which are an important part of the micro-level pathway of power, see especially Andrew Walter (chap. 3) and Louis Pauly (chap. 9 in this volume).
it does mean that in a given issue area or geographic relationship, power begins at home. First and foremost, policy makers must be free (or at least relatively free) to pursue national objectives in the specific issue area or relationship without outside constraint, to avoid compromises or sacrifices to accommodate the interests of others. Only then will a state be in a position, in addition, to enforce compliance elsewhere. Autonomy, the internal dimension, may not be sufficient to ensure a degree of foreign influence. But it is manifestly necessary—the essential precondition of influence.
The Core of Monetary Power
Autonomy, of course, is prized by governments in every aspect of international relations. Its salience, however, is most evident in economic relations, which by definition create a condition of interdependence with other states that is both active and ongoing. Economic relations involve transactional linkages, creating a web of mutual dependencies. Mutual dependencies, however—as Robert Keohane and Joseph Nye long ago reminded us in their classic Power and Interdependence, first published in 19773—are rarely symmetrical. Opportunities are created, therefore, for an exercise of influence by those who are less dependent—in short, by those with relatively greater autonomy. The lower the degree of a state's dependence on a relationship, relative to others, the greater will be its ability to manage existing connections to its own advantage.
And in no area of economic relations is the salience of autonomy more evident than in the realm of monetary affairs, where states are inescapably linked through the balance of payments. The risk of unsustainable payments disequilibrium repre- / sents a constant threat to policy independence. Excessive imbalances automatically generate mutual pressures to adjust, to help move the balance of payments back toward equilibrium. But adjustment can be inconvenient or even costly in both economic and political terms. No government likes being forced to compromise key policy goals for the sake of restoring external balance. All, if given a choice, would prefer to see others make the necessary sacrifices. At the macro-level of monetary affairs, therefore, monetary power consists of the capacity to avoid the burden of adjustment required by payments imbalance.
The core importance of autonomy in this regard has not always been fully appreciated in the scholarly literature. Indeed, most students of monetary power (including most of the contributors to this volume) prefer to stress the external dimension—the capacity to control the behavior of others in one way or another— rather than the internal dimension. But we cannot ignore the functionally derivative nature of the external dimension. In practice, power in a given issue area such as monetary relations logically begins with autonomy—the preservation of key policy goals at home. That is the necessary condition. Only if a state is actually able to avoid
3. RobertO. Keohane and Joseph S. Nye, Power and Interdependence: World Politics in Transition, 3rd ed. (New York: Longman, 2001).
34      Bmjamin J. Cohen
Macrofoundations of Monetary Power 35
the burden of adjustment domestically will it be in a position, in turn, to exert influence elsewhere. Hence, if we are interested in getting to the very core of power at the macro-level, we must go first to the internal dimension, as I propose here. Above all, what matters for the exercise of power abroad is practical freedom of action at home.4
The Two Modes of Influence
But we cannot ignore the external dimension entirely. Because monetary relations are inherently reciprocal, a potential for influence, in a real sense, is created automatically whenever practical policy independence is achieved. By definition, a capacity to avoid adj.i-mvenr costs implies that if payments equilibrium is to. be restored, others must adjust itis;v<i<l- at lo.isl part of tin burden will be divrru-d elsewhere. Hence, a measure of influence is necessarily generated as an inescapable corollary of the process. That too matters for analytical purposes.
But it is also important to keep the matter in perspective. The influence that derives automatically from a capacity to avoid adjustment costs represents at best a contingent aspect of power because it can be said to exist at all only because of the core dimension of autonomy. Moreover, the impacts involved are diffuse and undirected. That is very different from what is conventionally meant by the external dimension of power, which most often is understood to imply some degree of direct focus or deliberate intent—what Andrews (chap. 1 in this volume) calls a "purposeful act." From a political economy point of view, the difference is critical.
Essentially, the difference goes to the contrast between what Scott James and David Lake label the first and second faces of hegemony (or power): the first face of direct government-to-government influence, which is exercised through positive or negative sanctions; and the second face of market leverage, which favorably alters incentive structures.5 Correspondingly, we may think in terms of two modes in the exercise of influence: passive and active. The influence generated as a corollary of the adjustment process is exercised passively, even unpremeditatedly, and is best understood simply as the alter ego of autonomy. Alternatively, influence may be exercised actively, targeted at specific countries and applied with self-conscious purpose—in
4. Implicit in this formulation, of course, is an assumption that, in responding to external imbalance, the country is already at internal balance—that is, that domestic policy goals are already being achieved in spite of (or perhaps because of) the external imbalance. Hence, any compromise of (deviation from) current policy would in fact be considered a sacrifice. Implicit also is an assumption that domestic policy goals are denned by a political process rather than by pure economic calculation. No presumption is made that freedom of action at home will be used in a manner that meets the test of economic rationality.
5. Scott James and David Lake, "The Second Face of Hegemony: Britain's Repeal of the Corn Laws and the American Walker Tariff ofl846," International Organization 43, no. 1(1989): 1-30. In all, James and Lake identify three faces of hegemony, drawing, as they readily admit, on an earlier literature going back to Peter Bachrach and Morton Baratz, "The Two Faces of Power," American Political Science Review 56, no. 4 (December, 1962): 947- 52. The third face added by James and Lake, which stresses the hegemon's use of ideas and ideology to influence opinion, is more a part of the micro-level pathway of power and is not directly considered here.
} .   .       the language of Andrews (chap. 1 in this volume), a deliberate "influence attempt."
Both modes of influence begin with autonomy as a basic and necessary condition, I   . and in both cases other states may feel compelled to adjust. But, whereas in the pas-
sive mode the pressures exerted on others are market-driven, operating through E hegemony's second face, in the active mode the pressures are exerted directly by gov-
ernment, hegemony's first face. 1 .  . In a sense, passive influence in the adjustment process is relatively uncontrover-
i sial, broadly accepted as an unavoidable, if regrettable, consequence of inequality—
[ : a veritable fact of life. Active influence attempts, by contrast, are apt to become far , more politicized because they are both elective and purposeful. The active mode
[ seeks to compel others to bear the burden of adjustment, taking us well beyond the
: notion of influence as simply an incidental by-product of autonomy. The active
"•■ mode, in effect, aims to translate passive influence into practical control through the
■I' ■ instrumental use of power. That is a very big difference, indeed.
t '■ ■
' V The Two Hands of Monetary Power
The bottom line is clear. Whereas payments disequilibria are necessarily shared— one nation's deficit is someone else's surplus—the costs of adjustment need not be shared at all. Governments thus have every incentive, ceteris paribus, to maximize : their capacity to avoid adjustment costs—their autonomy—relative to others. The
;. greater the relative capacity to avoid adjustment costs, the greater is a state's mone-
■: ■'. . tary power.
; ■■ . My focus here on adjustment costs is hardly novel, of course. Other scholars have
i .. also placed the distribution of the burden of adjustment at the heart of their com-
ments on monetary power, including David Andrews,6 Randall Hertning,7 Jonathan ; Kirshner,s Michael Webb,9 and of course the late Susan Strange.10 But most treat-
ments until now have been regrettably ambiguous about what is meant by adjust-I . ment costs, leaving the analysis incomplete. We still lack a full understanding of
i what, precisely, the notion of burden is supposed to mean in the context of payments
.[: adjustment. Hence, we still lack a full understanding of the macrofoundations of
[. monetary power as well.
!.. . To help promote a fuller understanding, I propose to resurrect a distinction that
:: I first outlined in a much earlier attempt to explore the concept of adjustment
.[■ 6. David M. Andrews, "Capital Mobility and State Autonomy: Toward a Structural Theory of In-
:[ ternational Monetary Relations," International Studies Quarterly 38, no. 2 (1994): 193-218.
'[ 7. C. Randall Henning, "Systemic Conflict and Regional Monetary Integration: The Case of Eu-
■■irope," International Organization 52, no. 3 (1998): 537—73.
j ■. 8. Jonathan Kirshner, Currency and Coercion: The Political Economy of International Monetary Power
:i : (Princeton; Princeton University Press, 1995).
; 9. Michael C. Webb, "Capital Mobility and the Possibilities for International Policy Coordination,"
Policy Sciences 27, no. 4 (1994): 395-423.
: 10. See, especially, Susan Strange, States ami-Market, 2nd ed. (London: Pinter Publishers, 1994). For
an evaluation of Strange's thoughts on monetary power, see Cohen, "Money and Power in World Politics"; Eric Helieiner (chap. 4 in this volume).
36      Benjamin Cohen
Macrofoundations of Monetary Power 37
costs.11 Specifically, I distinguish between two distinctly different Irinds of adjustment com one consilium:.;, rhe other i'lrsiiional. Corresponding to each of the two kinds of adjustment cost is a very different kind of monetary power, which we may call the two "hands" of power,12 At the macro-level, monetary power is fundamentally dual in nature. On the one hand, states have the Power to Delay; on the other hand, they have the Power to Deflect. A two-fisted government prefers both.
The continuing cost of adjustment, we shall see, may be defined as the cost of the /j new payments equilibrium prevailing after all change has occurred. The Power to Delay is the capacity to avoid the continuing cost of adjustment by postponing the process of adjustment.
The transitional cost of adjustment, by contrast, may be defined as the cost of the ( change itself. When the process of adjustment cannot be put off, the Power to Deflect represents the capacity to avoid the transitional cost of adjustment by diverting as much as possible of that cost to others.
The Continuing Cost of Adjustment
To understand the Power to Delay, we must begin with the concept of adjustment. By definition, adjustment imposes on deficit countries a real economic loss that will persist indefinitely once the process is complete. This is the continuing cost of adjustment. Nothing suits the interest of deficit countries more than a capacity to postpone adjustment for as long as possible.
Payments Adjustment
The standard measure of balance in the balance of payments is the current account, which comprises all transactions relating to a country's current national income and expenditures—imports and exports of goods (merchandise trade) and services ("invisibles") plus unilateral transfers. Adjustment, correspondingly, is the process by which imbalances in the current account—surpluses or deficits—are reduced or eliminated. Import and/or export volumes adjust to restore payments equilibrium. Countries with deficits experience a decline of imports of goods and services relative to exports; countries with surpluses experience the reverse.
Not all imbalances need to be eliminated, of course. Standard economic theory teaches that many current-account imbalances are simply the result of what may be regarded as a kind of rational intertemporal trade—deficit countries borrowing resources from the rest of the world for productive investment at home and surplus countries investing savings abroad today to support greater domestic consumption tomorrow. Such imbalances, in principle, are sustainable indefinitely and require no
11. Cohen, Adjustment Costs:
12. My choice of the term hands here, while perhaps a bit whimsical, is intended to be consistent with the anatomical bent of the faces-of-power literature.
adjustment at all. In practice, however, many imbalances go well beyond what can
j be readily sustained for all kinds of reasons—for example, because borrowed funds
: are not invested productively or because of financial-market limitations. In such in-
j stances, which are all too frequent in the real world, adjustments of trade volumes
[ are indeed required.
i      . Adjustments of trade volumes, however, are impossible, without a corresponding
i  . reallocation of productive resources;13 and in a market setting, resource reallocations
j: will not occur without the stimulus of a change of prices or income. The required
| .'■ price and income changes may be promoted directly by means of so-called expen-
I diture-changing policies that aim to alter the overall level of spending, such as mon-
'!. : etary and fiscal policy; or they may be promoted more indirectly via a change of the
:j   : exchange rate—which in the traditional economics literature is referred to as an ex-
! penditure-switching policy, promoting adjustment via an altered ratio of prices be-
! tween tradable and nontradable production,14 Formally, adjustment may be defined
!'.: as "a marginal reallocation of productive resources and exchanges of goods and ser-
| vices under the influence of changes in relative prices, incomes, and exchange
1 ■' rates."15 This is the classical concept of "real" adjustment, the basic tool of open-
:j  . .: economy macroeconomics.
■j Real adjustment is necessarily a mutual process, reflecting the reciprocal nature of monetary relations. Just as one economy cannot be in deficit without others being in surplus, so resources cannot be reallocated in one without equivalent and offsetting
j ;' reallocations elsewhere. Should a deficit country move resources into export pro-
(■■'.' duction that were previously employed in producing for the home market, surplus
j countries will also find themselves obliged to shift resources about as they begin to
j .. receive additional imports. Likewise, should a deficit country increase output in import-competing industries, surplus countries will find themselves exporting less and
;| thus with additional resources for use in nontraded production. In either case, the
! reallocation of resources is complementary; the process of adjustment is shared.
Redistributing the Pie
However, although the process of adjustment is necessarily shared, the same need not be true of the burden of adjustment. In fact, once equilibrium is restored, the deficit country will unavoidably suffer a real economic loss, which will persist indefinitely. This is the continuing cost of adjustment, which is always borne wholly by deficit countries.
.5 ■
.] 13. Reallocation of resources should be understood to mean not only switches from one type of em-
ploytnent to another but also switches to or from unemployment, i 14. The efficacy of exchange-rate changes will depend, of course, on the extent of pass-through—
\ that is, the degree to which domestic tradable-goods prices actually move in response to changes of nom-
| inal currency values. In practice, pass-through may be limited or delayed considerably. Trade restrictions
; | ; or capital controls may also be regarded as expenditure-switching policies but are not directly considered
::i here.
15. Cohen, Adjustment Costs, 3.
38      Benjamin J. Cohen
Mam/foundations of Monetary Power
39
To comprehend why, assume a simple two-country model of payments imbalance. For the deficit country, adjustment requires a reduction of imports relative to exports, which is possible only if its real national absorption of goods and services, the sum total of spending by all domestic residents, is reduced relative to that of the surplus country. At the new payments equilibrium, therefore, the deficit country must be worse off than the surplus country, in the sense that it will now receive a smaller proportion of the combined output of the two economies. That is what I mean by the continuing cost of adjustment. I label it a continuing cost because it is open-ended—the ongoing sacrifice imposed by the new equilibrium that prevails after all change has occurred.
In absolute terms, the magnitude of the continuing cost may vary considerably, depending on the particulars of the approach to adjustment. The required change in the current account can be accomplished via a very different combination of changes in real national income and absorption in deficit countries—for example, a reduction of absorption relative to a more or less stable national income; an absolute loss of national income as well as absorption (via unemployment or an unfavorable movement of the terms of trade); an increase of national income, all of which, however, is absorbed abroad; or even an absolute increase of absorption as well as national income. Whatever the approach taken, however, the bottom line remains the same. At the new equilibrium, deficit countries will receive a smaller share of combined world output—a thinner slice of the pie. That is a sacrifice no matter how you cut it.
Deficit countries, therefore, have every incentive to put off the process of adjustment for as long as possible. Delay pays. As long as there is no change in the status quo, there will be no redistribution of the pie—hence no new burden. The scale of a state's Power to Delay is indicated by its capacity, in relative terms, to effectively postpone the payments adjustment process.
The Transitional Cost of Adjustment
But that is only one hand of monetary power. The continuing cost of adjustment involves an ongoing sacrifice imposed by the new equilibrium prevailing after all change has occurred, that is, after the adjustment process is concluded. But the process itself also imposes a sacrifice—the cost that must be incurred to make the necessary change. Each adjustment implies transition, a once-for-all phenomenon; and each transition has its own cost, separate and quite distinct from the presumed burden of the new equilibrium obtaining after the transition is complete. That is what I call the transitional cost of adjustment—in effect, the price of getting from here to there. Governments have every incentive to avoid this cost, too. No country wants to make more sacrifices than absolutely necessary.
The Adjustment Process
To illustrate the nature of the transitional cost of adjustment, consider a worker who, having lost a job and being unable to find a comparable one, finally accepts a
• lower-paying position. This process of adjustment imposes two costs on the worker.
! The more obvious one is the real sacrifice implied by the new position, namely, the
| difference between the new wage and the previous wage. This is an open-ended phe-
t. ..          namenon, a loss of income that will go on as long as the worker remains in the new
I position—the continuing cost of adjustment. But, in addition, the worker must have
I. suffered some loss of income during the period of enforced idleness. There may have
| been some real cost incurred in searching for a new job, investing in new skills, or
j moving to a new location. This is a once-for-all phenomenon, a singular loss of in-
'. come associated with the process of change itself. That is what I mean by the tran-
I sitional cost of adjustment.
j The question is, who pays? In the illustration, the burden falls on the worker. But this need not always be so. The government, for instance, might provide unemploy-
} ment compensation, job training, or other forms of adjustment assistance, thus
i shifting at least some of the cost to the taxpayer. Alternatively, part of the burden
; might be borne by the worker's former employer in the form of a generous sever-
! ance package or even by private charitable organizations dedicated to aiding the
\ involuntarily unemployed. In fact, the distribution of the transitional cost of ad-
i justment is, a priori, indeterminate. Unlike the continuing cost of adjustment, which
is never shared, the transitional cost is, in effect, up for grabs,
i Recall that the process of ba!ance-of-payments adjustment necessarily involves a realignment of relative prices, incomes, or exchange rates sufficient to generate the required reallocation of resources at the margin. The greater the changes of prices,
f incomes, or exchange rates required, the greater is the transitional cost of adjustment.
■I In principle, payments equilibrium can be restored either by real depreciation—
| policies of monetary deflation or nominal currency devaluation/depreciation—in
j deficit countries or by real appreciation—monetary inflation or nominal currency
i revaluation/appreciation—in surplus countries. Implications for the distribution
■ of the burden of adjustment differ greatly depending on which route is taken. Both
| economic and politica! elements of cost are involved.
i
I Fixed versus Floating Exchange Kates
j The circumstances under which this transition takes place matter, of course. Con-
sider first a world in which nominal exchange-rate changes are ostensibly ruled out—in today's terminology, a world of "hard" pegs. In that case, distributional implications are reasonably straightforward. With formal devaluations or revaluations ruled out, payments equilibrium can only be restored through expenditure-changing policies. That is, adjustment will be accomplished through either a market-j driven fall of prices and incomes in deficit countries, reinforced by restrictive mon-
| etary and fiscal policies, or a market-driven rise of prices and incomes in surplus
! countries, reinforced by more expansionary monetary and fiscal policies. In the
former case, it is plainly the deficit countries that bear the burden of adjustment. Economically, deflationary conditions will almost certainly result in higher unemployment, slower growth, and perhaps even recession before a new equilibrium can be established. Politically, austerity is bound to erode a government's popularity with
40      Benjamin J. Cohen
Macrofoundations of Monetary Power 41
voters. Conversely, in the latter case, it is the surplus countries that pay the price. Accelerated inflation reduces purchasing power and can distort investment incentives. It also tends to be politically unpopular.
Alternatively, consider a world of exchange-rate flexibility, where nominal exchange-rate changes are possible—in today's terminology, a world of "soft" pegs or some manner of floating. In this case, distributional implications are more complex because governments are no longer limited to expenditure-changing policies alone. Policy makers now can "pick their poison," as a recent International Monetary Fund (IMF) study puts it."" External adjustment can be allowed to impact prices and incomes in the domestic economy either directly with the nominal exchange rate fixed, indirectly via the expenditure-switching effect of exchange-rate movements, or via some combination of the two. In such a world, two separate aspects of the process are influential in determining the costs involved—one involving any movements of exchange rates that do occur, the other involving the degree of domestic price and income changes that ultimately are required, whether nominal exchange rates move or not.
Suppose some exchange-rate movements do occur as part of the adjustment process. Who bears the onus of responsibility? A realignment of rates may be the result of deliberate policy decisions (formal devaluation/revaluation) or may be essentially market driven (nominal depreciation/appreciation). Either way, governments may be held accountable for triggering or tolerating changes in a currency's nominal value.
Does this matter? In a hypothetical two-country world, where currency values are the inverse of one another, it should make no difference who is seen as responsible for the change. Exchange-rate movements would be symmetrical, a decline of one country's money necessarily equivalent to a rise of the other's. But in the real world of more than ISO currencies, by contrast, the distinction can matter a great deal. The evolution of a given money's value in relation to any other single currency, its bilateral exchange rate, may be substantially different from the evolution of its value against the population of currencies in general—called the effective exchange rate. A change in one money's effective exchange rate, even if sizable, may have little impact on individual bilateral rates if spread broadly enough. Conversely, even a small change in an effective exchange rate may have a very large impact elsewhere if it is concentrated on just one or two bilateral rates. In short, exchange-rate movements may be anything but symmetrical. As a practical matter, therefore, some governments may be exposed to much more criticism than others, even if they are not the first mover.
Essentially, this is a political issue. Exchange-rate changes are difficult to ignore. An exchange rate is like the eye of a needle through which prices of all domestic goods and services are linked and compared with the prices of foreign output. Because this
16. Shigeru Iwata and Evan Tanner, "Pick Your Poison: The Exchange Rate Regime and Capital Account Volatility in Emerging Markets," International Monetary Fund, Working Paper 03 /92 (Washington, D.C., 2003).
role makes the exchange rate a critical variable in determining the pattern of resource allocation as well as the level and distribution of income, governments have every reason to avoid the onus of responsibility insofar as possible. Nominal exchange-rate changes can generate a considerable backlash among voters, especially (but not exclusively) in small economies, for symbolic as well as material reasons. Devaluation or depreciation is typically interpreted as a defeat for a government's policies, damaging its reputation and credibility. Conversely, revaluation or appreciation may be resented for its potentially painful impacts on balance sheets and the earning capacity of key sectors of the economy. As a practical matter, few governments wish to be blamed for a sizable change in the value of the national currency.
The second aspect of the transition process in a world of floating rates concerns the degree of impact on the domestic economy. Once adjustment is under way, who experiences the greatest price or income changes? Governments may also be held accountable for any domestic austerity or inflation that results from the process of restoring external equilibrium.
, This matters because we know that domestic impacts, too—not just exchange-rate movements—may be anything but symmetrical. In practice, prices and incomes may change much more in some countries than in others, depending on circumstances. Adjustment in one country could generate relatively little macroeconomic change at home but considerable price and income pressures abroad, effectively diverting much of the pain of adjustment elsewhere; or, conversely, most of the impact could be bottled up domestically, whether exchange rates move or not. As with exchange-rate movements, few governments wish to be blamed for a sizable impact on the domestic economy.
Summary
Overall, then, the distribution of the transitional cost of adjustment depends on both aspects of the process: first, who bears the onus of responsibility for any exchange-rate changes that occur; and second—whether exchange rates change or not—who is forced to experience the biggest direct changes of domestic prices and income. In monetary affairs, these are the price of getting from here to there, which are also sacrifices no matter how you cut it. No wonder governments want to avoid the transitional cost of adjustment, too, deflecting as much as possible to others! The scale of a state's Power to Deflect is indicated by its capacity, in relative terms, to effectively divert the transitional cost of adjustment to others.
The Power to Delay
What, then, are the sources of monetary power at the macro-level? What are its limits? States obviously differ greatly in their relative capacity to avoid the burden of adjustment. It is equally obvious that there are limits to the autonomy of even the most powerful states. How can all this be explained?
42      Benjamin J, Cohen
Macrofoundations of Monetary Power 43
Given the dual nature of the macro-level pathway, it should not be surprising that separate factors might be at work in each of the two hands. Most critical for the Power to Delay, I suggest, are financial variables—above all, a country's international liquidity position, which encompasses both foreign reserves and access to external credit. The more liquidity there is at a country's disposal, relative to other states, the longer it can postpone adjustment of its balance of payments. Most critical for the Power to Deflect, by contrast, are more fundamental structural variables, also defined in relational terms, that determine how much real sacrifice will be required once the process of adjustment begins. It should also not be surprising that there might be distinctly different limits to each of the two hands of monetary power.
International Liquidity
A country's international liquidity comprises all available sources of internationally acceptable liquid assets. Before the postwar revival of global capital markets, the term was generally assumed to be synonymous with the sum of a country's international reserve assets. But once financial globalization began to take hold, the meaning of the term was expanded to include access to external credit as well, extended to the government or to the private sector. Today, international liquidity is generally defined to encompass the full array of international means of payment owned by or available to a country's public authorities and residents.
The ultimate purpose of international liquidity is financing: to cover deficits in the balance of payments, via either a net reduction of external claims (owned reserves) or a net increase of external liabilities (borrowing). The availability of financing to an economy, relative to others, can have a significant impact on the timing of adjustment and, hence, on the distribution of adjustment costs among deficit countries. More liquidity means more capacity to stave off any unwelcome reallocation of resources. Every deficit country has an obvious incentive to postpone the continuing cost of adjustment for as long as possible. The longer one deficit country can manage to put off adjustment, the greater will be the pressure on other deficit countries to bear the burden instead.
Of course, surplus countries too may have an incentive to delay the adjustment process—for example, if they believe that once the process begins, it is they who will be compelled to bear the bulk of the transitional cost of adjustment. Moreover, should that be their preference, surplus countries also have a greater ability to delay adjustment because it is almost always easier to absorb surpluses than to finance deficits. The motivation of surplus countries, however, is unlikely to be as intense as that of deficit countries, which have both costs to worry about. Moreover, even surplus states must anticipate the possibility that, sooner or later, they will suffer deficits, too. Hence, all states have a rational interest in acquiring and maintaining a healthy international liquidity position, on which the Power to Delay depends.
What, then, are the limits of this hand of monetary power? This requires a closer look at each of the two main components of international liquidity: owned reserves and borrowing capacity. The conditions affecting each are similar but not identical.
Owned Reserves
Superficially, it might seem that a government would want to hoard as many reserves as possible; insulation from payments pressures would be maximized by the largest possible stockpile of usable liquid assets. But that neglects the cost involved in acquiring reserves, which must be balanced against the benefit of greater auton-j omy. Reserves can be accumulated either as a result of current-account surpluses or
| by borrowing. Both strategies mean a reduction of real national absorption, either
| directly as a result of reduced imports relative to exports or indirectly as a result of
increased interest payments. Neither, therefore, is likely to be pursued without limit I because the cost of acquiring reserves could turn out to be greater than the loss of
\ absorption that might be required by adjustment. Economic theory has long argued
that rational policy makers can be expected to seek an optimal level of reserves rather than a maximum.
Optimality, however—like beauty—lies in the eye of the beholder. Different pol-i icy makers can make very different calculations, depending on their subjective eval-
\ uations of the costs and benefits involved. And these evaluations, in turn, will very
! much depend on politics, international as well as domestic. A government that feels
; beholden to constituencies that would be especially hurt by a reduction of payments
deficits, such as large-scale importers, would be likely to discount the cost of hoarding additional reserves. By contrast, a government that feels it can count on foreign '. allies to bail it out in the event of a payments emergency would be less inclined to
I invest in new reserves. A priori, therefore, no generalization is possible about where
the limits are likely to be found in this context. All we know for sure is that the ap-t .. petite for owned reserves will be considerably short of infinite. Hence, the Power to
t Delay by this means will be short of infinite, too.
j
Borrowing Capacity
In most respects, much the same also can be said about external borrowing. Here, too, it might appear that a government would want to make as much use as possible of borrowing capacity to finance deficits. The more liquidity that can be raised ex-j ternally, either by the government itself or by the private sector, the longer adjust-
\ ment can be postponed. But that too neglects the costs involved. These costs include
j not just the direct debt-service payments that would be required by foreign loans;
j even more critically, they include possible policy compromises that could become
j necessary if the country finds itself overextended to foreign creditors.
1 External credit can be raised from a variety of sources, of course. But whatever
the source, the liquidity provided can turn out to be too much of a good thing should the level of borrowing appear to rise beyond the economy's capacity to service the debt. For poorer and less developed countries, the main source of external credit is the public sector—governments of the more advanced industrial economies or multilateral agencies like the IME Overextension to public-sector creditors usually means that the borrower ends up negotiating a stabilization program, either bilater-
44       Benjamin^. Cohen
Macrofoundations of Monetary Poměr 45
ally with creditor governments, multilateraliy through the mechanisms of the so-called Paris Club, or with the IMF, with all the attendant conditionality. For middle-income emerging markets or more advanced economics, the main source of external credit is the global capital market. Overextension to private creditors usually means, eventually, a loss of perceived creditworthiness, which can lead to a sudden halt in new lending just when it might be most needed. Worse, excessive borrowing risks provoking panicky withdrawals and crisis, as capital importers around the world have sadly learned, from Mexico in 1994-95 to east Asia in 1997-98 to Argentina early in the new millennium. Reputation in financial markets, as we know, is a fragile flower, difficult to cultivate but easy to uproot. Painful policy adjustments may be required to restore a country's access to private investment.
Whatever the source of credit, therefore, autonomy may eventually have to be sacrificed for the sake of restoring external balance—a direct loss of power. Hence, with borrowing too, just as with owned reserves, rational policy makers can be expected to seek an optimum rather than a maximum. And here too, calculations of optimal-ity will very much depend on politics,
But there is also a big difference. The calculations demanded here are inherently more complex than they are with owned reserves because they necessarily involve tricky questions of probability and risk. With reserves, evaluations of prospective costs are relatively straightforward. Little risk is associated with hoarding reserves, and the real losses from deficit reduction or interest payments can be estimated with a reasonable degree of certainty. With external credit, by contrast, nothing is certain because borrowing capacity is by definition subjective in nature, often fluctuating widely, and even wildly, in response to the fickleness of creditor governments or changing sentiment in the marketplace. Because of this uncertainty, generalizations about limits are even more difficult than they are with the reserve component of liquidity. .
In effect, limits are not set by borrowers at all. Rather, they are set by creditors, both public and private. It is they who gain the power that overextended debtors lose. The challenge for borrowers is hard enough when dealing with creditor governments, whose decisions may be ruled as much by politics as economics. Calculations are even more difficult when it comes to market actors, who are constantly judging what they perceive as the quality of policy performance in individual economies. Financial markets are like a perpetual opinion poll. If a country is currently able to avoid a deficit reduction owing to ready access to credit, it is because the markets have given it their Good Housekeeping Seal of Approval. Conversely, if a country finds itself no longer able to put off an adjustment owing to a cessation of lending, it is the markets that are enforcing a limit on its Power to Delay. The more states relay on borrowing capacity rather than owned reserves for their international liquidity, the greater is the role of creditors, public and private, in determining who ultimately will be forced to undergo real adjustment.
Two implications follow. First, it seems clear that the distribution of the continuing cost of adjustment among deficit countries will be heavily influenced, if not largely determined, by creditor perceptions of debt-service capacity, which tend to
favor the relatively wealthy. Ceteris paribus, the Power to Delay should be greatest in the advanced industrial economies—the nations that enjoy the highest standing as international borrowers. The Power to Delay will he least in poorer and less developed economies that have limited access, at best, to foreign finance. Second, it also seems clear that the distribution of the continuing cost among deficit countries is apt to be highly volatile, given the persistent threat of rapid swings of sentiment about the "soundness" of policy in one economy or another. The perpetual opinion poll often changes its mind—and when it does, the ability to postpone adjustment through borrowing is changed as well. Taken together, these two observations suggest that, although wealthier economies may be the most favored in this context, there is no fixed pattern involved. What creditors giveth by way of a Power to Delay, they may also taketh away.
The Special Case of the United States
Finally, this brings us to the special case of the United States, with its unparalleled capacity to postpone adjustment. Since 1981, the U.S. current account has been in deficit in all but one year (the recession year of 1991)—a record unlike that of any other country. The United States clearly enjoys more Power to Delay than anyone else. How can this be explained?
The answer lies in the unique status of the dollar as the world's preeminent international currency—indeed, the world's only truly global currency. The United States enjoys the most Power to Delay because the greenback is "king of the world," as one journalist has put it, "the world's bedrock currency."17 Global popularity translates directly into a sustained demand for the dollar or dollar-denominated claims, which in turn enables the United States to finance deficits, in effect, with its own money. A need for international liquidity in the conventional sense is obviated when national liquidity is all that is required.
But there is also a downside to this privilege. Dollar accumulation around the world is no more than a form of external borrowing by the United States. In acquiring dollars or dollar-denominated claims, foreigners automatically extend credit to the U.S. economy; in the case of greenback notes, the credit is even interest-free. As with all external borrowing, therefore, there is a potential limit, set by the willingness of foreigners to go on lending. The ability of the United States to postpone adjustment ultimately rests on that same perpetual opinion poll, that is, on the judgments of agents elsewhere, including not only private-market actors using the dollar for investment purposes but also foreign central banks using the dollar for their reserves. Should the perpetual opinion poll lose its faith in the dollar—ceasing to lend or, worse, seeking to liquidate past investments—the United States could find itself under great pressure to reverse its current deficit. Today, many believe, the danger is even greater now that a potentially attractive alternative to the greenback
17. Rick Hampson, " Whatever They Think of America, People and Nations the World Over Prefer the Greenback," USA Today, December 26, 2001, p. A01.
46      Benjamin jf. Cohen
Macrofoundations of Monetary Power 47
is available in the form of Europe's new joint currency, the euro.18 The U.S. Power to Delay is by no means limitless.
The Power to Deflect
The Power to Deflect, by contrast, derives not from financial variables but, rather, from more fundamental structural variables that distinguish one national economy from another. Two features in particular stand out: the degree of openness and the jiegree of adaptability of each individual economy.
Some observers might wish to add a third feature: whether an economy happens to be in surplus or deficit. But that would be a mistake. Initial payments positions obviously are relevant to the distribution of the continuing cost of adjustment and therefore to the Power to Delay. But when it comes to the transitional cost of adjustment, as indicated, distribution is effectively up for grabs.
At issue, to repeat, are two questions. First, who bears the onus of responsibility for any exchange-rate changes that may occur? Second, whether exchange rates change or not, who is forced to experience the greatest direct changes of domestic prices and income? These are the two critical aspects of the adjustment process that bear on the distribution of the transitional cost. Each may fall on either surplus or deficit countries.
Sensitivity and Vulnerability
In an attempt to explore some of these issues in I966,191 suggested the notion of adjustment vulnerability, defined as the proportion of the transitional cost of adjustment borne by each economy. In essence, adjustment vulnerability might be understood as an inverse measure of what I here call the Power to Deflect. But I would not uše the term adjustment vulnerability today because it unfortunately obscures a now more familiar distinction, first introduced by Keohane and Nye in Power and Interdependence (1977), which helps us to understand why the two structural features of openness and adaptability, defined in relational terms, are of greatest salience in determining the Power to Deflect.
In exploring the nature of interdependence, Keohane and Nye broke ground in distinguishing between the two critical dimensions of sensitivity and vulnerability. Sensitivity interdependence, as Keohane and Nye put it, involves the susceptibility of an economy to impacts from the outside—the degree to which conditions in oné country are liable to be affected, positively or negatively, by events occurring elsewhere. Vulnerability, by contrast, involves the reversibility of impacts from the outside—the degree to which (or, in other words, the cost at which) a country is capable
18. Not everyone agrees with this assessment of the euro's potential attractiveness. For a more skeptical view, see Benjamin J. Cohen, "Global Currency Rivalry: Can the Euro Ever Challenge the Dollar?" Journal of Common Market Studies 41, no. 4 (2003): 575-95.
19. Cohen, Adjustment Costs.
of overriding or accommodating to the effects of events occurring elsewhere. The distinction is relevant here because it highlights the fact that every adjustment process can be decomposed into two separate elements: stimulus and response. The stimulus is the initial impact of disequilibrium on an economy; response refers to the ease with which the initial impact can be reversed. The sensitivity-vulnerability dichotomy neatly captures these two elements for analytical purposes.20
Openness and Adaptability
The Power to Deflect is a function of both elements of the adjustment process, stimulus and response. Openness matters to the Power to Deflect because it is the key determinant of an economy's sensitivity, relative to others, to payments disequilibrium (stimulus). Adaptability matters because it is the key determinant of an economy's relative vulnerability to disequilibrium (response).
Of these two structural variables, openness is clearly the easier to identify empirically. A standard measure of openness is the ratio of foreign trade to gross domestic product (GDP). The logic of its salience here is equally clear. The more open an economy, the greater is the range of sectors whose earning capacity and balance sheets will be directly impacted by adjustment once the process begins. This is true whether exchange rates remain pegged or are allowed to move. Either way, openness makes it difficult for an economy to avert at least some significant impact on prices and income at home.
In addition, if exchange rates move, governments in open economies are likely to come in for more criticism than would policy makers in more closed economies. Openness, ceteris paribus, also broadens the range of domestic constituencies that will take an active interest in the value of the country's currency. In a relatively closed economy, even fairly substantial exchange-rate movements may leave the largest part of the population unaffected and therefore indifferent, effectively insulating the government from criticism. In a more open economy, by contrast, where more interest groups will be directly affected, even small movements may lead to widespread opprobrium for policy makers, even if the government had nothing to do with starting the process in the first place. A high degree of openness makes it difficult to suppress widespread domestic repercussions when exchange rates change. It therefore makes it difficult for the authorities to deflect blame for any inflation or austerity that may result.
Adaptability is more difficult to identify empirically—it is an admittedly amorphous concept that, in fact, encompasses a myriad of qualities at the microeconomic
20. Much credit for recognizing the usefulness of the sensitivity-vulnerability dichotomy in this context goes to David Andrews, who has highlighted the notions of monetary sensitivity and vulnerability in a series of papers going back to "Bargaining Power and Policy Interdependence: Monetary Diplomacy in the Postwar International System," a paper presented at the 1997 annual meeting of the American Political Science Association. The point has also figured prominently in extended private correspondence between the two of us over many years. I am happy to acknowledge my intellectual debt to him on this point.
48      Benjamin J. Cohen
Macrofoundations of Monetary Power 49
level, such as factor mobility, informational availabilities, and managerial resilience. Still, the logic of its salience, too, is clear. For any given degree of openness, the adaptability of an economy determines how readily diverse sectors can reverse a disequilibrium without large or prolonged price or income changes. At issue is alloca-tive flexibility. The more easily productive resources can switch from one activity to another, overriding or accommodating to outside pressures, the less likely it is that domestic repercussions will involve serious pain; hence, the less likely it is, as well, that the process of adjustment will generate widespread resentment or protest. Conversely, the greater the rigidities characteristic of an economy's labor or product markets, the more serious will be resulting market dislocations and therefore the potential for political fallout. Adaptability, like beauty, may be one of those properties that is difficult to define, yet we know it when we see it and we know that it is important.
Implications
Again, two implications follow. First, it seems clear that the distribution of the transitional cost of adjustment is likely to favor larger and more diversified economies. Large size, as measured by GDP, generally means a relatively lower degree of openness. Greater diversification in production means that the economy offers more opportunities for alternative employment when adaptations are required. Smaller and less developed economies, conversely, are likely to be the least favored in the adjustment process. Some three decades ago, in the midst of the massive dislocations generated by the first oil shock, I wrote about what appeared to be a "cascading" of the burden of adjustment among oil-importing countries, with the poorest and least developed economies being forced to bear the greatest burden of all.21 "Power economics," I called it then. Today, with the wisdom of hindsight, I would call it a manifestation of the Power to Deflect.
The second implication is that, unlike the continuing cost of adjustment, the distribution of the transitional cost of adjustment can be expected to be comparatively stable over time rather than volatile. Structural variables such as openness or adaptability tend to change relatively slowly, to the extent that they change at all. The Power to Deflect, accordingly, is likely to change slowly, if at all, as well.
From Passive to Active Mode
Finally, we return to the measure of influence that is inherent in the Power to Deflect. Although the essence of the Power to Deflect is a capacity to avoid the transitional cost of adjustment (autonomy), the practical effect, as we have noted, is to divert the burden elsewhere, compelling others to bear it instead—a form of influence. In and of itself, the influence that is generated in this manner, which I have de-
21. Benjamin J. Cohen, "Mixing Oil and Money," in Oil, the Arab-Israel Dispute, and the Industrial World' Horizons of Crisis, ed. J. C. Hurewitz, 195—211 (Boulder: Westview Press, 1976).
scribed as the alter ego of autonomy, is passive and diffuse, essentially a product of market forces. But a more active mode is also possible, as many authors emphasize.22 The active mode, stressing the direct use of positive or negative sanctions in government-to-government relations, seeks to translate passive influence into practical control through the instrumental use of power. What is the connection between the two modes?
The connection, clearly, lies in the politics of interstate relations and, especially, in what Andrews (chap. 1 in this volume) calls monetary statecraft. The active mode—"influence attempts that rely primarily on the manipulation of monetary affairs," to quote Andrews—is optional. It is also purposeful, seeking to enforce compliance by way of pressure or coercion. In other words, it is policy-contingent. This means that it is not enough simply for a state to enjoy the structural characteristics essential to the Power to Deflect. Relative openness and adaptability are necessary conditions but hardly sufficient. One can think of a number of larger and more diversified economies that seem capable of diverting the transitional cost of adjustment to others, including, especially, the advanced industrial countries. But not many of these are known to engage in direct arm-twisting to get their way on monetary issues. Beyond a capacity for influence, a government must also have the motivation to put its Power to Deflect to active use—an understood framework of policy goals. Motivation will reflect a host of considerations peculiar to an individual country, involving foreign-policy strategy and domestic institutions as well as underlying constituency politics and political culture. As several of the contributions to this volume make clear, there is no certainty at all that the capabilities created by the Power to Deflect will be actively exploited.
Conclusion
To summarize, we may say that the macrofoundations of monetary power are best understand as being dual in nature. At the macro-level, monetary power is deploy-able with two hands: the Power to Delay, aimed at avoiding the continuing cost^if adjustment; and the Power to Deflect, aimed at avoiding the transitional cost of adjustment, The Power to Delay is largely a function of a country's international liquidity position relative to others, comprising both owned reserves and borrowing capacity. The Power to Deflect has its source in more fundamental structural variables: the relative degree of openness and adaptability of the national economy. The Power to Delay is limited only by the government's appetite for reserves and by the willingness of foreign agents to lend. The Power to Deflect is limited by the economy's underlying attributes and endowments.
Accordingly, it should he no surprise that states vary greatly in their monetary power, implying a systematic element of hierarchy in monetary relations. In fact,
22. Including most of the contributors to this volume; for a prominent exception, see Eric Hefleiner (chap. 4 in this volume).
50      Benjamin J, Cohen
monetary relations have always tended to be distinctly hierarchical, taking the shape of what I have elsewhere described as a Currency Pyramid—narrow at the peak, where one or a few countries dominate, and increasingly broad below.23 Ultimately, for all states, the issue is adjustment costs. Relative standing in the Currency Pyramid depends on the relative capacity to avoid the burden of payments adjustment, malting others pay instead.
Recently, David Lake challenged international relations theorists to pay more heed to the element of hierarchy in international relations.24 In the light of current scholarship, he argues, state sovereignty can no longer realistically be regarded as an absolute principle—quite the contrary, in fact. In his words: "Hierarchy is, and always has been, part of international relations. . . . Our theories of international relations would be improved by explicitly incorporating variations in hierarchy."25 The analysis here does just that for theories of monetary relations, incorporating variations in hierarchy by exploring the underlying sources of monetary power. The practical importance of the analysis lies in its identification of the key factors that determine the relative power of individual states, all of which are amenable to public policy to a greater or lesser extent.
The positioning of states in the Currency Pyramid directly reflects their access to both hands of monetary power. At the peak of the Pyramid is the United States, long acknowledged as the most powerful state in monetary affairs. The analysis in this essay suggests that the dominant position of the United States, which many describe as a hegemony, should be attributed to the country's unique combination of relevant capabilities—the special privilege that it enjoys in financing deficits, due to the global role of the dollar, as well as the notable adaptability of its domestic economy, which also happens to be relatively closed as compared with most other nations. Conversely, the lowly status of many poor developing nations would appear to relate directly to their lack of international liquidity as well as, typically, to the relatively high openness and low allocative flexibility of their economies. In between, rankings may be said to depend on how the key liquidity and structural factors stack up in each individual country. If governments wish to elevate their standing in the Currency Pyramid, it is these factors that must be addressed,
23. Benjamin J. Cohen, The Geography of Money (Ithaca: Cornell University Press, 1998); The Future of Money (Princeton: Princeton University Press, 2004-).
24. David A. Lake, 'The New Sovereignty in International Relations," International Studies Review 5, no. 3 (2003): 303-23.
25. Ibid., 304.
CHAPTER THREE
Domestic Sources of International Monetary Leadership
Andrew Walter
Currency leaders enjoy various forms of influence and power. The "exorbitant privilege" of currency leaders, above all the ability to finance external deficits by issuing IOUs and thereby to delay adjustment, has in particular received great attention in the literature. But what produces currency leaders? What, in other words, are the sources of this central aspect of international monetary power? In the preceding chapter, Benjamin Cohen outlines the macrofoundations of international monetary power—that is, the general characteristics of states that allow them to delay payment of the continuing costs of adjustment or to deflect the transitional costs thereof. Cohen locates the principal sources of the Power to Deflect in states' fundamental economic characteristics, in particular in their relative economic size and openness. He goes on to identify the primary sources of the Power to Delay in states' overall liquidity position (the sum of their foreign reserves and access to international credit).
In this chapter, I build on this contribution while drawing attention to other elements of the literature on monetary policy. I do so in order to argue that there are two additional prerequisites of international monetary leadership, having to do with domestic policies and institutional arrangements. First, currency leadership requires a relatively conservative monetary policy from the leader that is credibly embedded in its domestic political and economic institutions. This credible policy framework helps to produce willing followership on the part of the key audience, private market agents, as well as other national monetary authorities. Second, currency leadership also depends on a related set of institutional arrangements that facilitate the emergence of highly developed financial markets. I dis-
I thank Eric Hcilemcr, Randy Henninjr, Jerry Cohen, Jonathan Kirshner, Scott Cooper, Bob Hancke, and various other participants at the Florence conference for their very helpful comments on an earlier draft. Special thanks are due to David Andrews for his detailed comments. None are responsible for any remaining errors.
51