Frameworks for Monetary Stability
Chapter

12 Should an “Independent” Central Bank Control Foreign Exchange Policy?

Editor(s):
Carlo Cottarelli, and Tomás Baliño
Published Date:
December 1994
Share
  • ShareShare
Show Summary Details
Author(s)
CARLO COTTARELLI

I might dream of a day of final triumph of central banking; that is when central banks are so successful in achieving and maintaining price and financial stability that currencies will be freely interchangeable at stable exchange rates. Then, as in Europe today, we could discuss merging independent monetary authorities into a collective central bank designed to preserve and institutionalize a stable common currency.

That is not for my lifetime—nor for any of yours.

Paul A. Volcker (1990)

In recent years, many countries have approved legislation granting their central banks autonomy from political power. Within a statutory mandate to pursue price stability, these banks are free to formulate and implement monetary policy. Typically, they have also been freed from the burden of financing the public deficit, given the right to control the supply of base money, and shielded from external pressures by provisions stipulating that they cannot “seek or take instructions” from either governmental or other bodies.1 In essence, they have been made “independent” in the pursuit of price stability.

In the world of many currencies within which central banks are doomed to operate in our lifetimes, central bank independence depends on the constraints existing foreign exchange arrangements place on monetary policy. These constraints may be powerful enough to define the monetary policy stance completely, leaving no room for “independent” monetary decisions. At a very abstract level, this situation is easy to explain. Monetary policy (for instance, the setting of the nominal money supply) determines the value of domestic money in terms of domestic goods. Foreign exchange rate policy (for example, the setting of the nominal foreign exchange rate) determines the value of domestic money in terms of foreign goods. These two values of money must eventually coincide. In the absence of free capital mobility, the identity between monetary and exchange rate policies is likely to hold only in the very long run. But when capital movements are unrestricted—a condition toward which many countries are now moving—monetary and exchange rate policies become inextricably linked, even in the short run.

The reciprocal relationship between monetary and foreign exchange policies and, eventually, their identity, is generally accepted at the theoretical level but often ignored at the institutional level. Even the new central bank laws, which give these banks full control of monetary policy, allow the government to control major aspects of foreign exchange policy, such as the choice of the foreign exchange regime, the definition of parities, and the setting of guidelines for foreign market intervention.

This paper maintains that a central bank cannot be fully independent in the pursuit of price stability as long as the government controls foreign exchange policy. Thus, if a country has opted for central bank independence, exchange rate policy decisions (except for choosing the type of regime, but not the parities) should be made by the central bank. If a country wishes to limit central bank independence, the appropriate solution is to allow a formal government override of both monetary and exchange rate policy decisions, as has been done in New Zealand. This solution is preferable to the alternative—separating monetary and exchange rate policy responsibilities—because it avoids potential conflicts that could result from inconsistency among policy instruments.

The paper first reviews different “models” of exchange rate policy legislation, and notes that, in general, governments retain major responsibilities in this area. It then shows that monetary and exchange rate policies are not independent of each other and discusses how the separation of monetary and exchange rate policy responsibilities affects central bank independence. Next it addresses the arguments that have been used in favor of this separation, while the final part discusses the degree of latitude that independent central banks should be allowed in their foreign exchange responsibilities.

One caveat is necessary. The paper focuses on formal responsibility for exchange rate policy. Obviously, political and economic realities differ from formal legislative provisions, for several reasons. For example, the most important foreign exchange (as well as monetary policy) decisions, regardless of formal responsibilities, will (and usually should) be made after consultations between the government and the central bank. Nevertheless, the formal assignment of responsibilities is important, as conflicts may arise between different economic policy decision makers.

Foreign Exchange Policy Responsibilities

The conduct of foreign exchange policy can be broken into five major areas of responsibility, namely: (1) the choice of the exchange rate regime (e.g., fixed exchange rate, float, float within a band, and so forth); (2) the choice of the parities in a fixed regime or under a limited float; (3) the setting of “directives” within a managed float regime; (4) daily decisions on exchange rate interventions; and (5) actual execution of intervention policy. Table 1 describes how, in selected countries, these responsibilities are split between the government and the central bank.2Table 1 shows that while the execution of intervention decisions always falls to the central bank, other responsibilities are usually split according to four basic “models”:

Table 1.Foreign Exchange Policy Responsibilities1
RegimeParitiesDirectivesInterventionImplementation
Japan2GVGVGVGVCB
France3GVGVGVGVCB
United Kingdom3GVGVGVGVCB
European Union4GVGVGVCBCB
Hungary5GVGVGVCBCB
Canada6GVGVGVCBCB
France7GVGVGVCBCB
Chile8GVGVCBCBCB
United States9
Germany10GVGVCBCBCB
Sweden11CBCBCBCBCB
Switzerland12CBCBCBCBCB
Croatia13CBCBCBCBCB

The table indicates whether responsibility for the component of foreign exchange policy reported in each column is attributed mainly to the central bank (CB) or to the government (GV).

Funabashi (1988, pp. 93–104).

Annual Report of the Committee of EC Governors (1992, p. 47). The line referring to France is based on the legislation that was in force before the 1993 Central Bank Law (see footnote 8 below).

This line refers to the provisions included in the Maastricht Treaty regulating exchange policy responsibilities in Stage III of the Economic and Monetary Union (see Appendix).

National Bank of Hungary (1991, p. 16). The government has control of the exchange rate, but the central bank can change the rate within a 5 percent band.

Duguay and Poloz (1991, p. 41). Foreign reserves are owned by the Government and are administered by the central bank as fiscal agent.

Article 2 of the 1993 central bank law.

Castello-Branco (1990, pp. 10–11).

Destler and Henning (1989) and Dominguez and Frankel (1993). Both the Treasury and the Federal Reserve have the legal right to decide on foreign exchange intervention. However, according to Dominguez and Frankel (1993a, p. 48), in practice Treasury supremacy in this area is recognized, despite the fact that most of the foreign exchange reserves used for intervention are owned by the Federal Reserve. Destler and Henning (1989, p. 88) take a more balanced view, arguing that “virtually no intervention takes place without mutual consent.”

Kennedy (1991, pp. 36–37).

Hörngren and Westman-Mårteneson (1991, pp. 172–74). The Bank of Sweden can represent the country in international organizations (Article 13 of the central bank law).

The foreign exchange policy independence of the Bank of Switzerland is more a matter of tradition than of specific legislation.

Article 3 of the Law on the National Bank of Croatia, promulgated on November 4, 1992, stipulates that the National Bank of Croatia “shall independently establish the objectives of monetary and foreign exchange policy, as well as the measures within the framework of its rights and obligations, in order to realize the established objectives of the monetary and foreign exchange policy.”

The table indicates whether responsibility for the component of foreign exchange policy reported in each column is attributed mainly to the central bank (CB) or to the government (GV).

Funabashi (1988, pp. 93–104).

Annual Report of the Committee of EC Governors (1992, p. 47). The line referring to France is based on the legislation that was in force before the 1993 Central Bank Law (see footnote 8 below).

This line refers to the provisions included in the Maastricht Treaty regulating exchange policy responsibilities in Stage III of the Economic and Monetary Union (see Appendix).

National Bank of Hungary (1991, p. 16). The government has control of the exchange rate, but the central bank can change the rate within a 5 percent band.

Duguay and Poloz (1991, p. 41). Foreign reserves are owned by the Government and are administered by the central bank as fiscal agent.

Article 2 of the 1993 central bank law.

Castello-Branco (1990, pp. 10–11).

Destler and Henning (1989) and Dominguez and Frankel (1993). Both the Treasury and the Federal Reserve have the legal right to decide on foreign exchange intervention. However, according to Dominguez and Frankel (1993a, p. 48), in practice Treasury supremacy in this area is recognized, despite the fact that most of the foreign exchange reserves used for intervention are owned by the Federal Reserve. Destler and Henning (1989, p. 88) take a more balanced view, arguing that “virtually no intervention takes place without mutual consent.”

Kennedy (1991, pp. 36–37).

Hörngren and Westman-Mårteneson (1991, pp. 172–74). The Bank of Sweden can represent the country in international organizations (Article 13 of the central bank law).

The foreign exchange policy independence of the Bank of Switzerland is more a matter of tradition than of specific legislation.

Article 3 of the Law on the National Bank of Croatia, promulgated on November 4, 1992, stipulates that the National Bank of Croatia “shall independently establish the objectives of monetary and foreign exchange policy, as well as the measures within the framework of its rights and obligations, in order to realize the established objectives of the monetary and foreign exchange policy.”

The British-Japanese model: the government controls all remaining aspects of exchange rate policy, including daily intervention decisions;

The Maastricht Treaty model: the government is responsible for choosing the exchange rate regime (including parities); in the presence of floating exchange rates, it formulates exchange rate policy directives;3 and the central bank conducts intervention policy within the government directives;

The German model: the government controls the regime (including parities), but the central bank is independent within certain constraints; in particular, the government abstains from setting exchange rate policy directives under a floating regime; and

The Swedish model: the central bank is responsible for all aspects of foreign exchange policy.

These models differ in the extent to which the government controls the exchange rate. In the first model, government control is complete under a fixed regime. Under a floating regime, the government may also maintain complete control by setting exchange rate policy guidelines and using intervention policy to implement them. The second model is similar to the first if the government “directives” are tight enough, although the central bank retains control of day-to-day intervention decisions. In the third model, the government may give up its foreign exchange policy responsibilities by opting for a floating exchange rate regime.4 Only the fourth model, however, allows the central bank to control the foreign exchange rate under all circumstances.

Assigning control over exchange rate policy to the government is the natural choice when the government is responsible for monetary policy, as it is in the United Kingdom or was in France before the 1993 reform. However, government control is more difficult to maintain when responsibility for monetary policy is attributed to the central bank, as it is in the United States, Germany, and the European Union (EU) (under Stage Three of the economic and monetary union (EMU)). The reason is that the central bank cannot be truly independent if the exchange rate is controlled by the government, a conclusion that follows from the fact that the money supply and the exchange rate are not independent instruments in the long run and, in most cases, in the short run.

The Identity of Monetary and Foreign Exchange Policies

No Private Capital Movements

The relation between the money supply and the exchange rate can better be appreciated by focusing first on a world in which there are no private capital movements. In its simplest form, this world can be described by the following set of equations (Dornbusch and Giovannini, 1990):

Equation (1) states the equality between domestic prices P and foreign prices P* (converted into domestic currency at the foreign exchange rate e). Equation (2) is the quantity equation, where M is the stock of money, V is velocity, and A is aggregate demand. Equation (3) defines nominal income Y as the product between real output y and domestic prices P. Equation (4) states that the excess (shortfall) of aggregate demand over supply equals the current account deficit (surplus). The latter, in the absence of private capital flows and other sources of money creation, equals the change in money supply.

In the above model, y, V, and P* are taken as exogenous. Out of the remaining five variables, P, A, and Y are not controlled directly by the monetary authorities. The latter can set either M or e, but not both. The choice of M as the exogenous variable corresponds to a flexible exchange rate regime, in which e adjusts to equate aggregate supply and demand (and the current account is always in equilibrium).5 If, instead, e is exogenously fixed in a fixed exchange rate regime, then the money supply becomes endogenous. If aggregate demand and supply differ for any initial value of M, the current account disequilibrium will imply a change in M until equilibrium is restored.

Could the central bank perpetuate the disequilibrium through complete sterilization—i.e., by offsetting the liquidity increase (decrease) caused by a current account surplus (deficit) through changes in domestic credit—severing the link between the money stock and the exchange rate?6 Such a move would cause a permanent disequilibrium in the current account and, hence, an infinitely large gain or loss of foreign reserves (Frenkel and Mussa, 1985, pp. 697–98). Obviously, a permanent current account deficit would not be sustainable, as eventually the central bank would run out of reserves. Even a permanent surplus could be unsustainable. In this case, sterilization would involve the acquisition of foreign reserves in exchange for the sale of nonmonetary central bank liabilities,7 whose yield would eventually have to rise with respect to the yield of foreign assets, thus weakening the central bank profit and loss account.8 In the long run, the money supply and the exchange rate cannot be independent instruments.9

Capital Movements

The above conclusion could be challenged on two grounds. First, it applies only in the long run, leaving room for a temporary separation of monetary and exchange rate policies. Second, it does not take into account that money velocity (V) is not constant but depends on the nominal interest rate. Both these objections can, however, be overlooked in the presence of private capital movements.

Consider first credible fixed exchange rates. In this case, owing to the absence of exchange rate uncertainty, foreign and domestic assets are perfect substitutes and the domestic interest rate must equal foreign interest rates. Any attempt by the central bank to raise (reduce) the money supply and lower (raise) interest rates—thereby affecting money velocity—would immediately be offset by capital outflows (inflows). The only effect would be a shift in the composition of central bank assets, but without any change in total base money. The money supply would be entirely endogenous. Sterilization would not be possible, even in the short run, because it would involve an immediate and infinitely large shift in the central bank portfolio.10

Things are somewhat different if, as in the case of a managed float or not fully credible fixed exchange rate, exchange rates are uncertain. Suppose that the exchange rate is exogenously set by the government. However, either because the exchange rate is not officially frozen, or because the fixed parity is not credible, people attach some uncertainty to the expected exchange rate level. It has been argued that, in this case, sterilized intervention is theoretically a feasible option. Following the mean variance approach to portfolio choice (Dornbusch, 1983 and 1984), the differential between domestic and foreign interest rates can be written as

where i is the domestic interest rate, i* is the foreign interest rate, δ is expected devaluation, and r is the risk premium. The risk premium, which investors require as an incentive to shift their portfolios away from the minimum variance portfolio, is a function of the relative (world) supply of domestic with respect to foreign assets. By changing the composition of the world stock of assets, sterilized intervention would affect the interest rate differential.

Suppose, for example, that the central bank wanted to sterilize a capital inflow after tightening monetary policy. Through open market operations, the central bank could raise the relative supply of domestic (government) paper, which, given the imperfect substitute hypothesis, would validate the equilibrium increase in the interest rate differential. In this way, the central bank would be able to affect both the domestic interest rate and the money supply without changing the exchange rate (or expected depreciation).11 Thus, “… sterilized operations (if effective)” would allow “the monetary authority to independently pursue foreign exchange and money stock targets” (Dornbusch and Giovannini, 1990).

The practical relevance of the above mechanism, and thus the relevance of sterilized intervention, has been questioned. In a meanvariance context, the effect of relative supply on the interest rate differential is, even theoretically, presumed to be quite small.12 Empirical research has confirmed that sterilized intervention has, at most, transitory and relatively limited effects (Obstfeld, 1983, Rogoff, 1984, Obstfeld, 1990, and BIS, 1988). According to the Bundesbank: “…there are limits to what can be achieved by a pure intervention policy. The monetary crises under the Bretton Woods system showed that powerful market trends cannot be suppressed through exchange market interventions by central banks, and more recent monetary history has reaffirmed this” (Deutsche Bundesbank, 1989, p. 26).13 The conclusion that the money supply and the exchange rate are not independent instruments is reinforced when capital mobility is present: only one of the two variables can be exogenously determined.14

Implications for Central Bank Independence

Potential Conflict

The fact that the money supply and exchange rate are interdependent strongly affects central bank independence. A distinctive feature of an independent central bank is its responsibility for monetary policy—i.e., for determining the money supply. What happens if the government controls the exchange rate, either under a fixed exchange rate regime or in the presence of sufficiently tight exchange rate policy targets that are binding on the central bank? Because the money supply and exchange rate are not independent instruments, attributing their control to two independent policymakers may cause conflict between monetary and exchange rate targets, eventually forcing one of the two policymakers to give up its independence.15

Of course, setting up an independent central bank could give rise to other, similar conflicts—for instance, the potential conflict between monetary and fiscal targets brought to light by the “unpleasant monetarist arithmetic” of Sargent and Wallace (1981). However, in the present world of large capital movements, the inconsistency of monetary and exchange rate targets may emerge much sooner and more violently than the inconsistency between monetary and fiscal targets, which emerges only if the government finds issuing bonds to finance the deficit increasingly difficult. International financial market openness substantially widens the scope for bond financing and can exacerbate the conflict between inconsistent monetary and exchange rate policies.

Moreover, legislation aimed at establishing central bank independence has specifically addressed the potential conflict between monetary and fiscal targets by, for example, setting tight ceilings on central bank credits to the government. However, such legislation often ignores the potential conflict between monetary and exchange rate policies.

The degree of inconsistency between monetary and exchange rate policies depends on how tightly defined exchange rate policy is. In a fixed exchange rate system, the inconsistency will emerge more clearly than it will in a system in which the government sets only broad exchange rate policy guidelines or in a system of floating within a band (Svensson, 1992). However, the conflict will disappear only if the government opts for a free floating regime. But even in this case, the government can still exert pressure on the central bank by threatening to reinstate a more binding regime.

The Outcome of the Conflict

What actually happens when the central bank’s money supply targets and the government’s exchange rate targets are inconsistent? Does the government necessarily prevail, reducing the practical relevance of central bank independence?

Economic theory cannot say much on this point.16 If the battle is actually fought—i.e., if the government actually sets the exchange rate at a level inconsistent with the monetary target and sticks to that level—the central bank is likely to lose control of the money supply, since sterilized intervention will be ineffective. However, the government may eventually be forced to forsake its initial exchange rate target—for example, under the pressure of public opinion. In practice, the outcome of a conflict between the central bank and the government will depend on a number of factors, such as the personalities of the head of the central bank and the minister of finance, the central bank’s reputation and overall standing among other policymaking institutions, and the degree of public aversion to inflationary policies.

The German experience gives some idea of the limits that are placed on central bank independence when the government controls exchange rate policy. Bundesbank presidents (Emminger, 1978, Pohl, 1990, Schlesinger, 1992, and Tietmeyer, 1991) have always lamented the constraints imposed on monetary policy by the exchange rate policies of the German Federal Government, including those stemming from the Bretton Woods system and later, from the European Community’s monetary arrangements.17 In fact, recent econometric evidence confirms not only that foreign exchange constraints did affect Bundesbank monetary policy but that sterilized interventions were effective only in the very short run (von Hagen, 1989). In some cases, the Bundesbank prevailed, forcing the Government to abandon certain foreign exchange objectives. In the 1968–69 confrontation, for example, the Bundesbank supported—and got—a revaluation of the deutsche mark despite the Government’s opposition. Holtfrerich (1988) recalls:

Huge pressures on the mark’s foreign valuation had developed in 1968 from enormous current account surpluses, and, finally, from speculative capital inflows. An inflation was imminent. . . . The Bundesbank…had pleaded for a real revaluation. . . . Bonn, however, resisted. . . . Public opinion, however, had rather applauded the Bank in this dispute. . . . In 1969, massive imports of speculative capital continued. The Bundesbank again pleaded for a real revaluation of the Mark. The Bonn cabinet again refused on May 9, 1969, but was split over this issue exactly along party lines. . . . This dispute became the major economic issue in the campaign for federal elections in September 1969. The result was a government led by the Social Democrats and, therefore, a revaluation of the mark by 9.3 percent in October 1969. . . . (pp. 146–47).

In most other cases of conflict, however, the Federal Government prevailed. Thus, the Government revalued the deutsche mark in March 1961, over Bundesbank objections (Holtfrerich, 1988, p. 145); brought Germany into the exchange rate mechanism (ERM) in 1979, although the Bundesbank initially had doubts about the move (Kennedy, 1991, pp. 80–81, and Pöhl, 1990, p. 8); and forced the Bundesbank to accept a relatively high conversion rate for the East German mark during German unification (Schlesinger, 1992, p. 3).18

Undoubtedly, these are relatively isolated episodes. In most cases, the Federal Government refrained from using the exchange rate as a weapon against Bundesbank policies.19 These examples are, however, illustrative of the constraints that monetary policy faces when the government controls the exchange rate. They also show that when monetary and exchange rate policies are assigned to different authorities, the decision process can be slowed down (the delayed appreciation of 1969 is a case in point). Indeed, delays and policy inaction are often seen as the main drawbacks of separating responsibilities for monetary and exchange rate policy (Destler and Henning, 1989, p. 153, and Dobson, 1991).

Arguments Against Central Bank Control of Exchange Rates

The preceding sections have shown that when a government controls exchange rate policy the central bank may lose its independence in the implementation or the timing of monetary measures. Is it logical to conclude that if central bank independence is to be taken seriously, central banks must be allowed to control both monetary and exchange rate policy? Five arguments have been used to support government control over exchange rate policy. The first three point to the costs of assigning exchange rate policy to the central bank. The other two stress that under certain conditions, central bank independence may not be inconsistent with government control of exchange rate policy.

Foreign Exchange Policy and Central Bankers

One commonly held view is that exchange rate policy is “too serious a matter” to be left to central bankers, particularly in a democracy. For example, Volcker (1990) argues that:

Governments will not—practically they cannot—divorce themselves from decisions on appropriate exchange rate regimes or, beyond some limits, actual exchange rate levels (p. 17). . . . at the end clearly there are points at which central bank policy intersects with vital considerations of public policy generally, and I think the exchange rate is a prime case in point. . . . I don’t think we can escape that in a democracy (p. 21).

In the same vein, Witteveen (1992, p. 32), in commenting on the provision of the Maastricht Treaty that assigns control of foreign exchange policy to the European Council, argues: “a compromise of some kind is necessary in this area. . . . Greater autonomy for the European Central Bank in these fundamental decisions cannot be acceptable to governments.”20

While these views do not explain why exchange rate policy might be of “vital” interest in terms of overall public policy, it is legitimate to suppose that this interest arises from the real consequences of exchange rate policy.21 In an ideal world of fully flexible prices, changes in the nominal exchange rate and money supply would have only nominal effects. In the presence of sticky prices, however, nominal variables have real short-term effects (i.e., until prices adjust). In the short run, for example, a nominal appreciation may lead to a real appreciation, affecting output and employment. Assigning exchange rate policy to the central bank may therefore entail some costs, not only because it prevents exploiting the short-term trade-off between employment and inflation, but also because central banks may be interested primarily in maintaining price stability and so may be perceived as not paying enough attention to unemployment.22 Arguably, then, in a democracy the choice between unemployment and price targets should be left to political forces.

However, this argument does not hold in the present context, because, if anything, it is an argument against central bank independence, and not against assigning control of exchange rate policies to an independent central bank. If political forces have decided that the economy will benefit from central bank independence, then consistency requires that the central bank be endowed with the instruments that will allow it to exercise its independence.23 Some critics of central bank independence have even argued that assigning exchange rate policy to the government is tantamount to stating the central bank should not be independent: “We sympathize with the view that the exchange rate is too serious a matter to be left to the Central Bank. The unavoidable implication of that view is, however, that the Central Bank cannot be independent” (Buiter and Kletzer, 1990).

However, arguing for central bank control of both monetary and exchange rate policies does not imply that central bank independence should necessarily be unbounded. Rather, it implies that the same set of constraints should be applied to both monetary and exchange rate policy instruments. The degree of central bank independence is a political choice. But, again, central bank independence should not be limited by creating confrontational situations such as those that could result from the separation of monetary and exchange rate policy responsibilities. An explicit limit on central bank independence would be the preferred course. This thinking lies behind the “over-ride” clause included in the recent New Zealand central bank law. While the law allows the Government to issue directives to the central bank in the areas of both monetary and exchange rate policy, it also forces the Government to take responsibility for any inflationary consequences.24

Domestic Policy Coordination Costs

Destler and Henning (1989, p. 153) argue that having the government and central bank share responsibility for monetary and exchange rate policy improves policy coordination. This argument is not very convincing if applied to the coordination between monetary and exchange rate policies, as the best form of coordination could be obtained by attributing both policies to the same authority.25 With respect to the coordination of fiscal policy (or other policies, such as trade policies), on one side, and monetary-exchange rate policies, on the other side, there is no doubt that a problem of coordination exists. Once again, however, opting for an independent central bank necessarily involves some loss of governmental control in all aspects of policymaking and will require some extra effort in terms of coordination (Begg and others, 1991, p. 15). The cost of this loss of control will have to be evaluated when a decision is made to set up an independent central bank.

International Policy Coordination Costs

It has been argued that granting central banks full authority over exchange rate policies “runs counter to the national governments’ inherent responsibility for exchange rate arrangements through membership in the International Monetary Fund” (Burdekin, Wihlborg, and Willet, 1992, p. 244). More generally, it has been argued that representation at the international level (including participation in Group of Seven meetings) is something that belongs to the realm of foreign policy, and that only the government can handle efficiently.

To evaluate a system in which central banks play a more substantial role in global economic policy coordination, it is necessary to appraise the system in which monetary and exchange rate policy responsibilities have traditionally been separated. Dominguez and Frankel (1993) note the “schizophrenia” that characterized monetary and exchange rate policy coordination before the Plaza Agreement.26 This schizophrenia may impede effective international cooperation. For example, Dobson (1991) argues that the difference between success and failure in the work of the Group of Seven has depended heavily on the degree of agreement between the Bundesbank and Germany’s Federal Government.27

Involving central banks more directly in international coordination could prove helpful.28 Discussions could be held on ways of formalizing this participation, such as government and central bank participation in international meetings aimed at coordinating monetary and exchange rate policies, as well as in Fund and World Bank meetings. Such joint participation would anyway be needed to achieve coordination between monetary and exchange rate policies, on the one side, and other economic policies, on the other. It would not be inconsistent with assigning final responsibilities for monetary policy to independent central banks.

Guarantees of Central Bank Independence

It has been argued that government control of foreign exchange policy is not inconsistent with central bank independence if appropriate guarantees are in place. The Maastricht Treaty adopts this solution: political decisions affecting the exchange rate must be made unanimously by a “qualified majority” and must promote the goal of price stability.29 If sufficient constraints are imposed on a government that retains control of exchange rate policy, the central bank may be able to operate with a high degree of autonomy. Once again, however, if the purpose of those constraints is to prevent the government from using foreign exchange policies in a manner inconsistent with monetary policy, a simpler solution would be to assign responsibility for exchange rate policy directly to the central bank.30

The Case of Bilateral Pegs

The argument that the central bank loses its independence if the exchange rate is pegged rests on the hypothesis that foreign monetary policy is exogenous, as it is for small countries or countries with unilateral pegs. However, a country may enter into a bilateral agreement to peg its currency to that of an equally large economy. Under such circumstances, the implication is that both central banks are committed to intervention when necessary. Suppose that country A is the more inflationary of the two countries. If the government of country B enters into a bilateral pegging agreement with country A, will the central bank of country B lose its independence? Not necessarily. If the exchange rate of country A tends to depreciate, country A’s central bank will have to buy its own currency and sell that of country B, reducing the money supply in country A. The central bank of country B will do the opposite. Both central banks may try to sterilize their foreign exchange intervention, perpetuating the disequilibrium. However, as the foreign reserves of country A are limited, its monetary policy will eventually have to be tightened. Thus, the tighter monetary stance of country B will prevail. The practical relevance of this situation is limited for most countries, however, as it is based on the hypothesis that domestic monetary policy affects external monetary policy, something that would be true only for large countries or countries that are party to bilateral pegging.

Conclusions: One Instrument, One Authority

Monetary and exchange rate policies should be assigned to the same decision maker, primarily because they represent two aspects of the same instrument. If it is decided that an independent central bank should handle monetary policy, the bank should also be responsible for foreign exchange policy. Otherwise, it will not be truly independent in either the scope or timing of its actions (Begg and others, 1991;Giovannini, 1992; Neumann, 1991; Fratianni and von Hagen, 1993; and MacArthur, 1990). As discussed in the first part of the paper, the conclusion that monetary and exchange rate policy responsibility should be assigned to the same authority is particularly cogent under the condition of unrestricted capital movement toward which most countries are now tending.

Thus, continued government control of central bank policy in countries that have opted for central bank independence is, in general, a remnant of the past. Can such provisions be justified as means of finding a “middle ground” between complete government control and central bank independence? Probably not. Even if policymakers decide on political grounds that a central bank’s independence should be limited (over and beyond the limits implicit in its mandate and accountability), dividing the responsibility for monetary and exchange policy between authorities should be avoided, as such division could create conflict and result in policy inaction. In this case, a solution similar to New Zealand’s (an explicit government override of both monetary and exchange rate policy decisions) would be preferable.

How extensive should the responsibilities of the central bank be in the area of exchange rate policy? As suggested by Begg and others (1991) the choice of the regime could be left to governments. Economic theory has not proved definitively that either a fixed or a flexible exchange rate is preferable, so this choice can be based on political grounds.

An independent central bank should have full control of the exchange rate. Under a floating rate regime, it should control the formulation and implementation of intervention policies and should not be allowed to receive or seek exchange rate policy directives from the government.31 If a fixed rate regime is chosen (either at the international or individual country level), central bank independence requires that the choice of parities, the currency to which to peg, and the timing and amounts of realignment should be the central bank’s responsibilities.32

An alternative solution would be to assign control of the parities to the government and authorize the central bank “to cease intervening in the foreign exchange market in support of an exchange rate target if it is judged that continued intervention would jeopardize internal monetary control” (Burdekin, Wihlborg, and Willet, 1992, p. 244, and Pöhl, 1990, p. 8). Such a solution would probably only create more conflict, as it would force the central bank to explicitly challenge the government’s decisions. Thus, this solution could provide less independence than what this paper proposes as necessary.

Should central bank jurisdiction be extended to other aspects of external policy? While this question goes beyond the scope of this paper, a tentative guideline might be that the central bank should have jurisdiction over instruments affecting, in the long term, primarily nominal variables, while the government should be in charge of instruments related to real variables. This principle provides a rationale for assigning control over monetary and exchange rate policies and capital movements to the central bank.33 At the same time, it implies that the government should retain responsibility for other external policy measures, such as trade policies (tariffs and quotas, among others), whose introduction (removal) could affect the equilibrium real interest and exchange rates (e.g., Edwards, 1989).

Appendix I

The Genesis of Article 109 of the Maastricht Treaty

Article 109 of the Maastricht Treaty splits responsibilities for exchange rate policy between the European Central Bank (ECB) and the European Council.34 Given the attention that this article has received during the debate on central bank independence, it is worthwhile to review its content and genesis. Note that the Treaty refers to the European Community (EC) rather than the European Union (EU).

Responsibilities for the five components of exchange rate policy singled out in Table 1 are assigned as follows.

Exchange rate regime: the EC Council of Ministers will determine the exchange rate regime; more specifically, it will have the power to conclude “formal agreements on an exchange-rate system for the ecu in relation to non-Community currencies.” The Council can conclude such agreements “after consulting the ECB in an endeavor to reach a consensus consistent with the objective of price stability,” and the decision must be unanimous.

Change of parities: the Council will also decide on changes in parities and determine when the system should be abandoned. The decision can be made by a qualified majority according to the proposal of the European Commission or the ECB (again after consulting the ECB).

Exchange rate policy directives: by the vote of a qualified majority, the Council can “formulate general orientations for exchange-rate policy” These general orientations “shall be without prejudice to the primary objective … to maintain price stability.” Such decisions also require prior consultation with the ECB.

Formulation and implementation of intervention policy: the ECB will be responsible for conducting foreign exchange operations, subject to the “general orientations” set by the Council.35

In sum, the Council is responsible for choosing the regime (including setting and changing parities) and for formulating general orientations for exchange rate policy. The ECB is responsible for formulating and implementating exchange rate policy in line with those general orientations. The question of whether these provisions are consistent with monetary policy independence for the ECB has been discussed at length. It has been argued, for instance, that the ECB will be protected by the required consultations and that the Council will be constrained by the objective price stability. It has also been noted that the decision to enter into exchange rate agreements must be unanimous. On the other hand, skeptics have stressed that it is not clear who will decide whether exchange rate policy decisions are consistent with the pursuit of price stability and that the consultations are not sufficient to guarantee central bank independence. Most likely, only future economic developments can determine the degree of independence the ECB will have.

It is worth stressing that Article 109 is the outcome of a compromise whose genesis can easily be traced back in the documents leading to the Maastricht Treaty. A separation of monetary and exchange rate policy responsibilities was already envisaged in the Werner Report, which assigned the choice of the regimes and parities to a “political” center, and of monetary policy and intervention policy to central banks (Baer and Padoa-Schioppa, 1989). The Delors Report reiterated this suggestion but also introduced the concept of the foreign exchange “directives” that would guide intervention policy in a floating exchange rate regime. More specifically, while arguing that intervention “would be made on the sole responsibility of the ESCB,” the Delors Report added that intervention would have to be “in accordance with Community exchange rate policy” (paragraph 60), thus implicitly stating that that policy would not be formulated by the European System of Central Banks (ESCB).

This point—namely the degree of exchange policy independence in a floating rate system—was heavily debated. The Monetary Committee of the European Community (1990) recognized the fact that members agreed the choice of the regime and parities should be left to the EC political authority. However, it also noted the conflict between the majority view, which argued that strategic decisions on exchange rate policy should be made at the political level, and a minority view, which supported full central bank independence in a floating rate regime (paragraphs 8 and 28–33). A disagreement along the same lines emerges in the initial (August 1990) draft of the Treaty prepared by the European Commission (1991, paragraph 2.4), and also in the November 1990 report of the Committee of Central Bank Governors (point f) putting forward the draft statutes of the ECB.36

While these official documents do not detail the composition of the “majority” and “minority,” the Bundesbank was clearly included in the latter. As argued by Tietmeyer (1991, p. 184): “… the ESCB can only be successful if it retains full control over the instruments that are necessary for conducting monetary policy. These include instruments not only for domestic monetary policy, but also for external monetary policy. Hence, the ESCB must also be given sole responsibility for exchange market intervention against currencies outside the system.”37

While the Bundesbank’s view was eventually defeated, the constraints placed on the ECB in the conduct of intervention policy were milder than those initially envisaged.38 Thus, a compromise solution eventually prevailed and, according to some (Thygessen, 1991, p. 478), made the ECB more independent in terms of intervention policy than the U.S. Federal Reserve or the Bank of Japan—but possibly not as independent as the Bundesbank.39

References

    AlogoskoufisGeorge and RichardPortes“International Costs and Benefits from EMU,”Chapter 5 in “The Economics of EMU” in European Economy Special Edition No. 1 (Luxembourg: Commission of the European Communities1991).

    BaerGünther and TommasoPadoa-Schioppa“The Werner Report Revisited,” in Collection of Papers Annexed to the Report on Economic and Monetary Union (Luxembourg: Commission of the European CommunitiesApril1989).

    Bank for International Settlements“Exchange Market Management and Monetary Policy,”inExchange Market Intervention and Monetary Policy (Basle: Bank for International Settlements1988).

    BeggDavid and others “European Monetary Union—The Macro Issues,”Part I of Monitoring European Integration: The Making of a Monetary Union (London: Center for Economic Policy Research1991) pp. 368.

    BruniFranco and DonatoMasciandaro“Evaluating Central Bank Independence: Theoretical Issues and European Perspectives,”Rivista Internazionale di Scienze SocialiVol. 99 (January-March1991) pp. 93118.

    BuiterWillemH. and Kenneth M.KletzerReflections on the Fiscal Implications of a Common Currency CEPR Discussion Paper Series No. 418 (London: Center for Economic Policy ResearchMay1990).

    BurdekinRichardC.K.ClasWihlborg and Thomas D.Willet“A Monetary Constitution Case for an Independent European Central Bank,”The World EconomyVol. 15 (March1992) pp. 2349.

    Castello-BrancoMartaCentral Bank Independence: The Chilean Case (unpublished Washington: International Monetary FundSeptember1990).

    Commission of the European Communities (1990a) Economic and Monetary Union Beyond Stage I (unpublished Brussels: Monetary Committee of the European CommunityJuly1990).

    Commission of the European Communities (1990b) “External Dimensions,”in “One Market One Money” European Economy No. 44 (October1990).

    Commission of the European Communities“Economic and Monetary Union,”Bulletin of the European Communities Supplement 2 (1991) pp. 562. Draft Treaty Amending the Treaty Establishing the European Economic Community with a View to Achieving Economic and Monetary Union.

    Commission of the European CommunitiesTreaty on European Union Text of the Treaty signed in MaastrichtFebruary71992 (Luxembourg1992).

    Committee of Governors of EC Central BanksAnnual Report1992.

    CrawfordMalcolmOne Money for Europe? (London: Macmillan1993).

    CukiermanAlexCentral Bank Strategy Credibility and Independence: Theory and Evidence (Cambridge, Massachusetts: MIT Press1992).

    DestlerI.M. and C. RandallHenningDollar Politics: Exchange Rate Policy-Making in the United States (Washington: Institute for International Economics1989).

    DeutscheBundesbankThe Deutsche Bundesbank: Its Monetary Policy Instruments and Functions Deutsche Bundesbank Special Series No. 7 3rd ed. (Frankfurt-am-Main: Deutsche Bundesbank1989).

    DobsonW.Economic Policy Coordination: Requiem or Prologue? Policy Analysis in International Economics No. 30 (Washington: Institute for International Economics1991).

    DominguezKathrynM. and Jeffrey A.Frankel (1993a) Does Foreign Exchange Intervention Work? (Washington: Institute for International Economics1993).

    DominguezKathrynM. and Jeffrey A.Frankel (1993b) “Does Foreign-Exchange Intervention Matter? The Portfolio Effect,”American Economie ReviewVol. 83No. 6 (December1993) pp. 135669.

    DornbuschRudiger“Exchange Rate Risk and the Macroeconomics of Exchange Rate Determination,”inThe Internationalization of Financial Markets and National Economic Policyed. by G.HawkinsR.M.Levich and C.WWihlborg (Greenwich: JAI Press1983).

    DornbuschRudiger“Comment,”inExchange Rate Theory and Practiceed. by John F.O.Bilson and Richard C.Martson (Chicago: University of Chicago Press1984).

    DornbuschRudiger and AlbertoGiovannini“Monetary Policy in the Open Economy,”inHandbook of Monetary Economicsed. by BenjaminFriedman and FrankHahn (Amsterdam: North-Holland1990).

    DuguayPierre and StephenPoloz“The Monetary Policy Decision-Making Process in Canada,”inThe Orientation of Monetary Policy and the Monetary Policy Decision-Making Process (Basle: Bank for International Settlements1991).

    EdwardsSebastianReal Exchange Rates Devaluation and Adjustment: Exchange Rate Policy in Developing Countries (Cambridge, Massachusetts: MIT Press1989).

    EmmingerOtmar“30 Jahre Deutsche Mark,”Monatsberichte der Deutschen Bundesbank (June1978).

    FrankelJeffreyA.“The Implications of Mean-Variance Optimization for Four Questions in International Macroeconomics,”Journal of International Money and FinanceVol. 5Supplement (March1986) pp. S5375.

    FratianniMichele and Jürgen vonHagen“European Monetary Union and Central Bank Independence,”Regional Science and Urban EconomicsVol. 23 (July1993) pp. 40125.

    FrenkelJacobA. and Michael L.Mussa“Asset Markets, Exchange Rates, and the Balance of Payments”inHandbook of International Economicsed. by Ronald W.Jones and Peter B.Kenen (Amsterdam: North-Holland1985).

    FunabashiYoichiManaging the Dollar: From the Plaza to the Louvre (Washington: Institute for International Economics1988).

    GiovanniniAlbertoCentral Banking in a Monetary Union: Reflections on the Proposed Statute of the European Central Bank CEPR Occasional Paper No. 9 (London: Center for Economic Policy ResearchAugust1992).

    GoodhartCharles“The Draft Statute of the European System of Central Banks: A Commentary,”Cahiers économiques et monétaires.No. 40 (1992) pp. 199224.

    GoodhartCharles“The External Dimension of EMU,”Recherches Economiques de LouvainVol. 59No. 1–2 (1993) pp. 6580.

    GrosDaniel and NielsThygessenEuropean Monetary Integration: From the European Monetary System to the European Monetary Union (London: Longman1992).

    HendersonDaleW.“Exchange Market Intervention Operations: Their Role in Financial Policy and Their Effects,”inExchange Rate Theory and Practiceed. by John F.O.Bilson and Richard S.Marston (Chicago: University of Chicago Press1984).

    HoffmanJ. and G.KeatingThe European System of Central Banks (London: Crédit Suisse and First BostonDecember1990).

    HoltfrerichCarl-Ludwig“Relations Between Monetary Authorities and Government Institutions: The Case of Germany from the 19th Century to the Present,”inCentral Banks’ Independence in Historic Perspectiveed. by GiovanniToniolo (Berlin: Walter de Gruyter1988).

    HörngrenLars and AnnWestman-Mårtensson“Swedish Monetary Policy: Institutions, Targets, and Instruments,”inThe Orientation of Monetary Policy and the Monetary Policy Decision-Making Process (Basle: Bank for International Settlements1991).

    HungJuannH. (1991a) Noise Trading and the Effectiveness of Sterilized Foreign Exchange Intervention Federal Reserve Bank of New York Research Paper No. 9111 (March1991).

    HungJuannH. (1991b) The Effectiveness of Sterilized U.S. Foreign Exchange Intervention: An Empirical Study Based on the Noise Trading Approach Federal Reserve Bank of New York Research Paper No. 9118 (May1991).

    KenenPeterB.“EMU After Maastricht,”Chapter 1 in EMU and ESCB After Maastricht Financial Markets Group London School of Economics (Washington: Group of ThirtySeptember1992).

    KennedyEllenThe Bundesbank: Germany’s Central Bank in the International Monetary System (London: Pinter Publishers for the Royal Institute of International Affairs1991).

    MacArthurAlanMonetary Operations Financial Market Development and Central Bank Independence (unpublished, Washington: International Monetary FundSeptember1990).

    MussaMichaelThe Role of Official Intervention (Washington: Group of Thirty1981).

    National Bank of Hungary“The Law on the Central Bank,”Market LetterNo. 10–11 (1991) pp. 18.

    NeumannManfredJ.M.“Central Bank Independence as a Prerequisite of Price Stability,”inEuropean Economy Special Edition No. 1 (Luxembourg: Commission of the European Communities1991).

    ObstfeldMaurice“Exchange-Rates, Inflation, and the Sterilization Problem: Germany, 1975–1981”European Economic ReviewVol. 21 (March-April1983) pp. 16189.

    ObstfeldMaurice“The Effectiveness of Foreign Exchange Intervention: Recent Experience, 1985–88,”inInternational Policy Coordination and Exchange Rate Fluctuationsed. by W.BransonJeffreyFrenkel and MorrisGoldstein (Chicago: University of Chicago Press1990).

    PöhlKarlOtto“Prospects of the European Monetary Union,”inBritain & EMU by KarlOtto Pöhl (London: Center for Economic Performance, London School of Economics1990).

    RogoffKenneth“On the Effect of Sterilized Intervention: An Analysis of Weekly Data,”Journal of Monetary EconomicsVol. 14 (September1984) pp. 13350.

    SargentThomasJ. and NeilWallace“Some Unpleasant Monetarist Arithmetic,”Federal Reserve Bank of Minneapolis Quarterly ReviewVol. 9(Winter1985) pp. 1531.

    SchlesingerHelmut“The Importance of Central Bank Independence for German and European Economic Policy,”Auszüge Aus Presseartikeln (Frankfurt-am-Main: Deutsche BundesbankDecember151992).

    SvenssonLarsE.O.Why Exchange Rate Bands? Monetary Independence in Spite of Fixed Exchange Rates CEPR Discussion Paper Series No. 742 (London: Center for Economic Policy Research1992).

    SwinburneMark and MartaCastello-Branco“Central Bank Independence and Central Bank Functions,”inThe Evolving Role of Central Banksed. by PatrickDownes and RezaVaez-Zadeh (Washington: International Monetary Fund1991) pp. 41444.

    ThygesenNiels“Monetary Management in a Monetary Union,”European Economic ReviewVol. 35 (1991) pp. 47483.

    TietmeyerHans“The Role of an Independent Central Bank in Europe,”inThe Evolving Role of Central Banksed. by PatrickDownes and RezaVaez-Zadeh (Washington: International Monetary Fund1991) pp. 17689.

    VolckerPaulA.“Triumph of Central Banking?”The 1990 Per Jacobson Lecture Central BankingVol. 2 (Autumn1991) pp. 7786.

    von HagenJürgen“Monetary Targeting with Exchange Rate Constraints: The Bundesbank in the 1980s,”Federal Reserve Bank of St. Louis ReviewVol. 71 (September/October1989) pp. 5369.

    WitteveenH.Johannes“Interview,”Central BankingVol 3 (Autumn1992) pp. 3138.

    WyploszCharles“Capital Controls and Balance of Payments Crises,”Journal of International Money and FinanceVol. 5 (1986) pp. 16779.

This paper has benefited from the comments of William Alexander, Lorenzo Bini Smaghi, José De Gregorio. Jorge Guardia, Manuel Guitián, Alain Ize, Gian Maria Mitesi-Ferretti, and Johannes Mueller.

See, for example, Article 7 of the Protocol on the Statute of the European System of Central Banks (ESCB) and of the European Central Bank (ECB), published by the Commission of the European Communities (1992).

The countries included in the group were selected according to the availability of reliable information on the distribution of foreign exchange responsibilities and may not reflect the relative frequency of different models around the world. For example, countries in which exchange rate responsibilities are assigned entirely to the central bank (at present, still an unusual feature) are probably overrepresented in the sample.

Strictly speaking, the European Council can set “general orientations” in the area of foreign exchange policy. This controversial aspect of the Maastricht Treaty is discussed in the Appendix.

As indicated in the table, the United States has been included in this third group, although the U.S. Government has direct responsibility for foreign exchange intervention. Within a floating exchange regime, the U.S. Government cannot formally set exchange rate policy guidelines that would restrain the independence of the central bank. Indeed, as discussed under Implications for Central Bank Independence, mere control of intervention may not significantly constrain the central bank in the absence of binding exchange rate policy guidelines.

In this case, ∆M = 0 and equation (4) is replaced by A = Y.

In this case, equation (4) would be replaced by A − Y = DCE, where DCE is domestic credit expansion, for ∆M = 0.

Initially, the central bank could sell government paper, but eventually it would have to resort to issuing central bank paper.

This would necessarily occur if central bank liabilities were not perfect substitutes with respect to other domestic assets—a clear possibility, as the former (being backed by foreign assets) are risk free, while other domestic assets are subject to default risk.

This result has been illustrated tor a model in which there are only tradable goods. In the presence of nontradables. the same conclusion applies as long as the real exchange rate (that is, the relative price of tradable and nontradable goods) is unaffected by the money supply. This hypothesis, which most economists are likely to accept as a long-term condition, is weaker than the assumption that the real exchange rate is constant.

Note also that, as the central bank cannot change domestic interest rates, it cannot affect money velocity. Thus, with capital movements, it is not necessary to assume (as it is in the discussion on the identity of monetary and exchange policies) that money velocity is constant.

It must be observed that this effect holds only if sterilized intervention involves outside assets—e.g., government debt, if the latter is considered as private wealth. Moreover, it requires that central bank assets are not considered part of private wealth (Henderson, 1984, p. 360).

Frankel (1986) shows that the effect of changes in relative asset supplies on the risk premium is equal to the covariance of asset returns times the coefficient of risk aversion, and the former is the square of a small fraction.

There are, however, cases in which sterilized intervention may be relevant. The first occurs when it signals future changes in the money supply (Mussa, 1981), the second when it is aimed at “pricking an existing speculative bubble” (Dominguez and Frankel, 1993a, p. 136: see also Hung, 1991a, 1991b). However, sterilized intervention cannot be used in either of these cases to affect the real exchange rate permanently. In the first case, the exchange rate would be affected only if the money supply were eventually changed. In the second case, intervention would work only insofar as it eliminated an existing disequilibrium of the real exchange rate. It must be recognized that the debate on the effectiveness of sterilized intervention is still very much active. In a recent paper, Dominguez and Frankel (1993b) argued that the “portfolio effect” is empirically relevant and allows for some separation of monetary and exchange rate policies.

In the presence of constraints on capital movements (e.g., a tax on capitai inflows or a quantitative restriction) a wedge can be created between domestic and foreign interest rates, thus freeing monetary policy even in the presence of fixed exchange rates. Capital constraints are. however, distortionary and tend to be circumvented eventually. In any case, even if capital constraints worked, they would not prevent the current account from deteriorating because of expansionary monetary policies used in the presence of fixed exchange rates. The mechanism described in the discussion on the identity of monetary and exchange policies would still be at work, forcing a realignment between the exchange rate and money supply in the long run (Wyplosz, 1986).

It is worth stressing that in a floating exchange rate system, government control of foreign exchange intervention policy is not sufficient to reduce central bank independence. Indeed, the discussion on foreign exchange policy responsibilities highlighted the fact that sterilized intervention cannot separate the exchange rate and money supply. Thus, if the central bank sterilizes changes in the money supply that arise from intervention, the intervention policy eventually becomes unsustainable and may lead to an infinitely large increase or loss of foreign reserves. The central bank has a problem when its money supply policy must be formulated according to a certain exchange rate target set by the government.

Cukierman (1992) discusses cases of conflict between the central bank and the government, but only for a closed economy.

Tietmeyer (1991, p. 180). for example, notes that “serious conflicts arose between the Bundesbank and the Federal Government over external monetary policy, especially under the old parity regime with the U.S. dollar at its center.” Pöhl (1990, p. 8) argues: “The conflict between ‘domestic’ and ‘external’stability is woven into the fabric of German monetary policy like a scarlet thread. . . . In my view, it would therefore be desirable to embody in the statute of a European central bank a clause to the effect that domestic stability must have priority over exchange rate stability.”

Admittedly, the latter example is, strictly speaking, not a case of exchange rate policy. Indeed, the term used in Germany to refer to the conversion of East German marks into deutsche mark was Umtauschverhältnis (conversion rate), rather than Wechselkurs (exchange rate). Nevertheless, the substance of the problem and its monetary implications were similar to those related to exchange rate policy

In part, this restraint may be an acknowledgment of the fact that sustaining aggregate demand through a devaluation may upset partner countries. For example, during the well-known conflict between Chancellor Helmut Schmidt and the Bundesbank in 1981, the Chancellor was “furious about the Bank’s restrictive monetary policy in a situation of alarmingly high and rising unemployment” (Holtfrerich, 1988, p. 148). However, it was clear that a devaluation within the ERM was out of the question, as economic crisis was affecting all EC countries. On that occasion, the German Chancellor, together with the French Government, resorted to a foreign loan to finance expansionary fiscal measures.

Tietmeyer (1991, p. 184) also argues that “decisions on the exchange rate regime, including the fixing of parities and central rates and their changes, will have to remain in the hands of the political authorities.” Finally, note that the provision assigning control of foreign exchange policy to the Council of Ministers was included in the draft European Central Bank statute prepared by the Committee of Central Bank Governors (Hoffman and Keating, 1990). The number of central bankers supporting government control of exchange rate policy is somewhat surprising. But, in Charles Goodhart’s words: “I doubt, however, whether the question is what the Governors “want,” rather what they think is practicable politics for their own countries” (Goodhart, 1992, p. 212).

For example, the Commission of the European Communities (1991) argued that “because the exchange rate is an important element of external economic policy, external monetary relations have a double aspect, monetary in a narrow sense and economic. They therefore require the participation of two actors, the monetary authority and the body responsible for economic policy” (p. 21). The EC Monetary Committee also argued that “exchange rates and intervention have implications for general economic policy, as well as for the Community’s balance of payments, responsibility for which falls partly on the political authorities” (Commission of the European Communities, 1990a, paragraph 31).

For example, fearing the inflationary consequences of a depreciation, the central bank may oppose such a measure, even when it is called for by a change in fundamentals (e.g., a worsening of the terms of trade). The result would be a forced adjustment through a decline in nominal prices, which in the short run would be accompanied by a recession.

It is worthwhile to recall that opting for central bank independence is a recognition of the fact that the long-run benefits of lower inflation are higher than the short-run costs arising from imperfect price flexibility.

Goodhart (1992, pp. 211–12, and 1993, p. 67) shares the view that an explicit override is preferable to decoupling monetary and exchange rate policy responsibilities.

Moreover, statutory provisions requesting “coordination” between a monetary policy controlled by the central bank and a foreign exchange policy controlled by the government (Swinburne and Castello-Branco, 1991) do not seem to be sufficient to guarantee independence.

Dominguez and Frankel (1993, pp. 51–52) seem to argue that such a “schizophrenia” persisted even after the Plaza Agreement. However, the role central bank governors have played at Group of Seven meetings since the agreement’s implementation has increased substantially—a tacit recognition of the fact that monetary and exchange rate policies must be effectively coordinated.

See on this point also Kenen (1992, pp. 103–104). The European Commission recognizes the existence of a problem of consistency between monetary and exchange rate policies but dismisses it by simply assuming that consistency is somehow achieved (Commission of the European Communities, 1990b, p. 190). The issue of representation for the EC at Group of Seven meetings after the setting up of the Economic and Monetary Union (EMU) is also discussed by Alogoskoufis and Portes (1991).

Of course, it is here assumed that coordination improves social welfare, a proposition which has also been subject to some criticism.

See Goodhart (1993, p. 66). Gros and Thygesen (1992, p. 417), and Crawford (1993, p. 207). Thygesen (1991. p. 479) also argues that: “Despite the precautions taken, this [i.e.. exchange rate policy] is the area where the greater risk to the ESCB’s autonomy and orientation towards price stability lies.”

With respect to the constraints introduced in the Maastricht Treaty, it has been noted that “price stability” is not easy to define. Even more difficult may be deciding whether specific decisions on that exchange rate are consistent with price stability. Unless this assessment is left to the central bank, however, this solution may not be very effective in protecting central bank independence (see Goodhart, 1993, p. 66, and Kenen, 1992, pp. 29 and 106).

Of course, monetary policy directives should also be prohibited.

While the use of foreign exchange reserves should be controlled by the central bank, formal ownership of foreign reserves could be maintained by the government. Even at present, ownership of and control over reserves have been separated in many countries, including in the United States. Severing the link between ownership and control is implicit in the choice of an independent central bank. Indeed, the government is likely to own the capital of an independent central bank but should be restrained by legislation from interfering with the bank’s policy decisions

As discussed, capital constraints can affect real variables, but, most likely, not in the long run.

The provision setting out these responsibilities is usually referred to as Art. 109 of the Treaty, and this convention will also be followed here. However, the relevant provision of the Treaty is paragraph 25 of Art. G. This paragraph redrafts, among other things. Art. 109 of the Treaty establishing the European Economic Community (the so-called Treaty of Rome). So, in principle, reference should be made to Art. 109 of the Treaty of Rome, as amended by the Maastricht Treaty.

This provision is actually included in Art. 105.

See also Bruni and Masciandaro (1991) referring to the views of the Report on EMU prepared by Commissioner Christophersen.

See also Hoffman and Keating (1990).

Art. 108 of the draft Treaty proposed by the August 1990 communication of the Commission stated: “The Council, acting by a qualified majority on a proposal from the Commission and in close cooperation with the European Central Bank, shall lay down directives for the Community’s exchange-rate policy. In accordance with those guidelines, the European Central Bank shall conduct an appropriate intervention policy”. In Art. 109 of the Maastricht Treaty, the term “directives” is replaced by the milder “general orientations.” formal consultations with the ECB are required, and the Community’s exchange rate policies are limited by the mandate to pursue price stability.

The issue of whether the Bundesbank or the ECB is more independent is, of course, debatable. For example, it must be recognized that both the monetary and the exchange rate policies of the European Union will be constrained, at least in principle, by the objective of price stability. The mandate of the Bundesbank is, to some extent, less precisely defined (“safeguarding the currency” is the expression used in Art. 3 of the Bundesbank statute), and there is no formal constraint on exchange rate policy decisions taken by the German Federal Government.

    Other Resources Citing This Publication