Frameworks for Monetary Stability

5 The ERM Experience

Carlo Cottarelli, and Tomás Baliño
Published Date:
December 1994
  • ShareShare
Show Summary Details

Continental European countries have had a long-standing preference for stable exchange rates that goes far beyond arguments related to the choice of nominal anchor. The inter-war traumas of competitive devaluation, protectionism, and unemployment explain some of the concerns about the use of the exchange rate as an active instrument of economic policy. Exchange rate fluctuations have also been seen in Europe as an important impediment to the postwar drive to integrate European markets, a sentiment reinforced by the renewed emphasis on market integration since the launching of the Single Market Program in the mid-1980s. More generally, exchange rate stability has been regarded as an essential building block for monetary union, the drive for which has a political as well as an economic basis.

In the early period of reconstruction after World War II, the emphasis in the monetary area was on restoring currency convertibility in Europe. As the realization of this goal neared, broader issues of monetary integration began to move to the fore. Already in 1962, the Commission of the European Communities (EC) was proposing concrete stages that would lead to a monetary union. With the Bretton Woods system of fixed exchange rates firmly in place, however, there was little urgency in the quest for a separate monetary system within Europe. As the U.S. balance of payments deficits of the 1960s began to undermine the Bretton Woods system, the need for a European alternative strengthened. The Hague summit of 1969 advocated the creation of a European Monetary Union, and the 1970 Werner Report proposed a three-stage approach to its attainment.

These ambitious proposals were soon overtaken by the collapse of the Bretton Woods system. Many in Europe were uncomfortable with the 2.25 percent fluctuation margins on either side of the central U.S. dollar parities that had emerged under the Smithsonian Agreement of December 1971. The EC countries opted in early 1972 for fluctuation margins among themselves that were half those against the dollar, thus forming the so-called “snake-in-the-tunnel.” With the emergence of generalized floating in 1973, the European currencies found themselves in a floating snake. The French franc, the Italian lira, the Irish pound, and the pound sterling had relatively short-lived membership in the snake; in effect, the snake became a limited deutsche mark zone, comprising the Benelux countries and Denmark, with Norway and Sweden as associate members for most of its existence.1

Following the initiative of German Chancellor Schmidt and French President Giscard D’Estaing, the European Monetary System (EMS) was founded in March 1979, building on the framework established by the snake. Its goal was to create a “zone of monetary stability in Europe.” Each EMS currency was given a central exchange rate in terms of domestic currency per European Currency Unit (ECU), from which could be deduced a series of bilateral central rates.2 The focal point of the EMS was its exchange rate mechanism (ERM): the central banks of countries participating in the ERM agreed to maintain their bilateral exchange rates within a fluctuation margin of 2.25 percent on either side of the bilateral central rate.3 Central rates could only be altered by mutual agreement.

Under the agreement, when two currencies reach a bilateral margin, both central banks are required to intervene without limit to defend the parity. The central bank of the weaker currency may borrow unlimited amounts of the stronger currency from the latter’s central bank, through the Very Short Term Financing Facility (VSTF) and then sell it on the foreign exchange markets to buy quantities of the weaker currency. In this way the money supply of the stronger currency expands while that of the weaker currency automatically contracts.4 This apparent symmetry was reinforced by the creation of a divergence indicator based on a currency’s divergence from its central ECU rate. When a currency reached 75 percent of its maximum possible divergence, there was a presumption that it would respond with offsetting policy measures.

The German Anchor Role in the ERM

It is clear that the ERM emerged more out of the general sentiment in favor of exchange rate stability than specifically as a means of controlling inflation in Europe. As it was formally symmetric and without an anchor, it did not lend itself easily to this latter task. Of course, Germany never intended the ERM to be a symmetric system and, in practice, the ERM proved far from symmetric.5 The deutsche mark, however, could only play an anchoring role if the other countries came to accept this (indeed, seek it). In the early 1980s, governments began to focus on the need to rein in inflation, which had been running at high levels after the accommodative policies pursued following two oil shocks of the 1970s. A consensus began to emerge that monetary policy should be directed at price stability, typically interpreted as maintaining a low and stable rate of inflation, along with a general acceptance of the need for a clear framework for achieving this goal. The role of the ERM as a device for transferring German anti-inflation credibility to the other countries was thus one that evolved and was not by design (at least not explicitly so).

The emergence of the widespread use of the exchange rate anchor was influenced by the existence of a framework—the ERM—that could easily accommodate it. But it is also clear that in many respects the circumstances in Europe seem particularly suitable for using the exchange rate as a nominal anchor. There is a large country, Germany, with a strong track record in keeping inflation low. The openness of the European economies implies an important direct link between exchange rate changes and inflation. Moreover, the strong trade links within Europe mean that a high percentage of producers see themselves directly or indirectly in competition with German producers, intensifying pressures toward price and inflation convergence.

Evolution of the ERM6

The ERM has evolved through five stages since its inception in 1979. The first phase, from March 1979 to March 1983, was marked by seven realignments. During this period there was little convergence in performance, and there were important differences in economic objectives across countries.

The beginning of the second phase is usually dated to March 1983 with France’s decision to abandon the policies of demand expansion and nationalization that had been pursued following the change of government in 1981.7 The devaluation of the French franc by 8 percent against the deutsche mark in March 1983 was accompanied by an adjustment program of monetary and fiscal austerity measures, supported by a loan of ECU 4 billion from the European Community. The goals of the program were to reduce inflation to competitor levels, to return the trade account to balance, and to stabilize the budget deficit.

During this second phase, which lasted almost four years, the role of the deutsche mark as nominal anchor came to the fore, as Germany’s ERM partners strove to bring down their inflation rates and to maintain their bilateral exchange rate pegs to the deutsche mark (Chart 1). There were only four realignments, and two of these involved one currency: the Italian lira in July 1985 and the Irish pound in August 1986.

Chart 1.Original ERM Currencies: Inflation and Exchange Rates

Source: International Monetary Fund, Research Department, Surveillance Data Bank.

The realignment of January 1987 marked the commencement of the third phase of the ERM or the “new” ERM as it was described by Giavazzi and Spaventa (1990). During this period the ERM was strengthened by the Basle-Nyborg agreement in 19878 and the drive for economic and monetary union (EMU) gathered momentum, reinforcing the desire to maintain stable exchange rates. Indeed, exchange rates became essentially fixed.9 The growing stature of the ERM was reflected in its expanding influence: Spain (1989), the United Kingdom (1990), and Portugal (1992) became members and four non-member currencies were linked to the ECU—those of Norway (1990), Finland (1991), Sweden (1991), and Cyprus (1992).10

During this period, the convergence of inflation rates continued to such an extent that German inflation was no longer the lowest in the ERM. However, beneath the facade of tranquility, the seeds of the later turmoil were growing. Outside what has become known as the “core,” sufficient nominal convergence had not yet been achieved to support unchanged exchange rates, and significant misalignments began to emerge. Second, the focus of German monetary policy was on combating the rise in inflation following unification; this resulted in interest rates that were rather high, given domestic conditions in many of the low-inflation ERM countries, especially as they moved into recession in 1992–93.

The underlying tensions came to the fore in early September 1992, and the next eleven months—which constitute phase four of the ERM—were marked by episodes of considerable turmoil in European foreign exchanges and sharp movements in short-term interest rates (Chart 2). In the period from September 1992 to May 1993, the Italian lira and the pound sterling withdrew from the ERM, the Spanish peseta was devalued three times, the Portuguese escudo twice, and the Irish pound once. Outside of the ERM, the Finnish markka, the Swedish krona, and the Norwegian krone unlinked their currencies from the ECU, with the first two experiencing significant effective depreciations. These developments corrected the exchange rate misalignments that had emerged over the preceding years, but the tensions resulting from the interaction of weak cyclical conditions and high real interest rates remained. With the markets seeing policymakers keen to reduce interest rates, and uncertain about the pace of interest rate reduction in Germany, enormous bets were placed in the course of July 1993 that the French and Danish currencies would be devalued relative to the deutsche mark. This speculation subsequently spread to the Belgian franc. In face of this massive assault, a decision was made to widen the intervention margins to 15 percent, effective August 2, 1993.

Chart 2.ERM Countries: Interest Rate Differentials1

Source: International Monetary Fund, Research Department, Surveillance Data Bank.

1Data tor the Netherlands are not included, as interest differentials against the deutsche mark have been relatively stable.

The decision to widen the intervention margins represented the beginning of the fifth phase of the ERM. At one level, this decision was aimed at containing the immediate crisis. However, it was also widely seen as a means of allowing European countries to cut interest rates where this was appropriate for their domestic circumstances, while allowing the Bundesbank to set a path of interest rates consistent with a return to price stability in Germany.

At first, exchange rates against the deutsche mark depreciated in anticipation that ERM countries would tolerate some weakening of their currencies relative to the deutsche mark to facilitate lower interest rates. Instead, in a number of countries, short-term interest rates, which had been raised to resist the speculation against the narrow bands, were kept significantly above pre-crisis levels, with a view to limiting currency depreciation. The heightened uncertainty was reflected in unusually large bid/ask spreads in the money markets. After a few months, the currencies that had weakened began to recover against the deutsche mark, and central banks in question allowed their short-term interest rates to fall back toward German levels.

Disinflation in the ERM

Since the inception of the ERM, both the level and the dispersion of inflation has fallen in participating member states. The considerable progress made toward the attainment of nominal convergence suggests that the use of the exchange rate as a nominal anchor has been a successful strategy. Rigorous assessment of the role of the ERM, however, is made difficult by ignorance of what inflation performance would have been if the ERM had never existed, or what it might have been under some alternative nominal anchor. It is interesting to note that inflation also came down notably in the other industrial countries, both in Europe and outside Europe (Chart 3).

Chart 3.Inflation: Three-Year Moving Average

(In percent)

Key: ERM: Belgium, Denmark, France, Ireland, Italy, Luxembourg, and Netherlands.

Non-ERM Europe: Austria, Finland, Norway, Portugal, Spain, Sweden, Switzerland, and United Kingdom.

Non-Europe OECD: Australia, Canada, Japan, New Zealand, and United States.

In assessing the role of the ERM, a key question is the extent to which it limited the short-run costs of reducing inflation. In particular, advocates of the ERM as a credibility-enhancing device would expect to see an improvement in the short-run unemployment/inflation trade-off for the ERM countries reflecting the gains in inflation-fighting credibility conferred by adherence to the ERM margins. Evidence in this regard has been mixed. Giavazzi and Giovannini (1989) found that statistical models of wage and price inflation tended to overpredict the actual rate of inflation for France from late 1982 onward, and for Italy from late 1986 onward. These results are consistent with a downward shift to inflation expectations in the countries concerned around the time that they strengthened their commitments to the exchange rate target. On the other hand, de Grauwe (1989), comparing the experience of ERM countries with other OECD countries, found a more pronounced worsening of the inflation/unemployment trade-off among the ERM countries when compared to the rest of the OECD, particularly after 1984.11 He concluded that the ERM forced countries to adopt a gradualist approach to disinflation. He also argued that countries such as the United Kingdom and the United States, with floating rates, could adopt a more radical approach to reducing inflation, by allowing the exchange rate to appreciate, that might ultimately have lower output cost.12 More recently, Egebo and Englander (1992) summarized their survey of the empirical evidence of ERM credibility effects as providing “little basis from which to conclude that adhering to the ERM bands has altered the trade-off between inflation and unemployment.”

All of these comparisons are impeded by important methodological difficulties. For example, in assessing the relative experience of ERM and non-ERM countries, or in looking for shifts within a country in the short-run unemployment/inflation trade-off, it may be difficult to control for any changes in the level of structural unemployment. Nor is it clear that an approach adopted in one country would be politically feasible in another; if the exchange rate anchor is politically more feasible, then it may be the best choice, even if technically it is not the most efficient.

Finally, some argue that the exchange rate anchor, by limiting the scope for discretionary action, may be more effective in promoting an institutional structure that will act as a lasting buffer against inflation. In particular, it may have had a particularly important role in marshaling support for the independence of central banks in many European countries and thereby moving monetary policy toward a more medium-term perspective.13 Whereas the exchange rate anchor inherently limits the role of the central bank, it is argued that a monetary policy framework that provides a more involved role of the central bank may be less likely to secure support for a curtailment of political influence.

Origins of the Recent ERM Turmoil

The details of the exchange market turbulence in 1992–93 and its causes have been extensively discussed elsewhere.14 The fundamental problem, as noted earlier, was that there had been insufficient reconciliation of the policies being followed by individual countries with the needs of the system. In part, this reflected the fact that inflation convergence had not progressed sufficiently in all countries to support the degree of exchange rate stability that countries sought after 1987. While measuring exchange rate misalignments is far from a science, some sense of the scale of the disequilibria is provided by the appreciation of real exchange rates in Italy, Portugal, Spain, and the United Kingdom in the late 1980s and early 1990s and the subsequent corrections (see Chart 4).

Chart 4.Real Effective Exchange Rates

(Based on consumer price index) (1987=100)

Source: International Monetary Fund, Research Department, Surveillance Data Bank.

Even after the realignments of late 1992 and the first half of 1993, there remained the more general problem of tensions between diverging cyclical positions and the convergent monetary policies required by the ERM. These were associated largely with the after-effects of inadequate policy coordination following German unification—in particular, the failure to contain fiscal imbalances in Germany and the reluctance of Germany’s ERM partners to consider a realignment. With monetary policy in Germany being kept tight to control inflation there, and its nominal interest rates setting a floor for the other countries, real short-term interest rates were high throughout the ERM, particularly in countries where inflation had fallen well below that of Germany. The problem was compounded by the lack of maneuver on the fiscal front and the high level of structural unemployment, which reflected the failure of policies to take advantage of strong growth in the late 1980s to make progress in these areas.

What is less amenable to economic analysis, however, is the timing of the recent exchange market turmoil. Exchange rate misalignments and unbalanced policy mixes can be sustained as long as policymakers are willing to bear the cost and this is credible to the markets. The experience in Europe over the past few years has also underlined that market skepticism of the sustainability of an exchange rate policy does not necessarily entail immediate market pressure for a change. This skepticism may fester and be unleashed by some triggering events that signify to the markets that policymakers may no longer be able or willing to resist a change in the exchange rate. In other cases, however, some significant events may trigger a change in market sentiment where questions had not existed previously.

To illustrate the above, one can point to some of the factors that were especially relevant in the recent ERM turmoil. First, the goal of EMU had been important in underpinning exchange rate stability in the late 1980s and early 1990s. When doubts emerged as to the ratification of the Treaty on European Union (the Maastricht Treaty) in the summer of 1992, markets began to count less on its convergence requirements as a means of ensuring discipline in those countries that had not yet converged. In countries where inflation performance was already satisfactory and the level of the exchange rate was not in question in a medium-term context, the ratification uncertainties raised questions as to whether authorities would still judge it worthwhile to endure high real interest rates as the recession in Europe deepened and as inflation in Germany proved unexpectedly difficult to subdue. Second, domestic political factors had similar effects in a number of countries. In addition to the pervasive political uncertainties in Italy, one can note the way dates in the political calendar (in particular, referendums or elections) at times focused the attention of markets, for example, in France and Spain. Finally, doubts of the durability of the ERM itself undermined the markets’ confidence in the medium-term strategy, for example, in Belgium.

Issues Arising out of the ERM Crisis

A Return to Narrow Bands?

It has become commonplace to note that the ERM had evolved prematurely into a system of fixed exchange rates in the period 1987–92 and that, until full nominal convergence is achieved, the ERM needs to be operated in a more flexible manner. It is tempting to move from this observation to seeing the recent crisis in the EMS as resulting from an unfortunate confluence of an exceptional temporary disturbance (German unification) and the inadequate attention that had been given over a number of years to the accumulation of cost and price divergences among ERM countries. In this interpretation, one could argue that the threat to the system is largely over and the way is open to a return to narrow bands: exchange rates now appear to be more in line with fundamentals, cyclical divergences are narrowing and German interest rates are on their way down to levels that are more in tune with macroeconomic conditions in the ERM as a whole. Care would have to be taken not to allow a significant accumulation of cost and price divergences in the future, but the continued progress on inflation convergence in 1993 would seem to augur well in this respect. Real shocks could, of course, be a disruptive force in the future, but a shock of the scale of German unification is unlikely to be repeated, particularly not in the anchor country.15

Taking a broader perspective, however, it is clear that the environment within which the ERM operates has changed considerably since the 1980s. Three differences are particularly relevant. First, the gathering momentum of the drive for EMU has resulted in even greater weight being attached to the goal of stabilizing exchange rates. Inconsistencies in policies and macroeconomic performance across countries, which were accommodated in a system in which parities changed frequently, become more exposed when parities are relatively stable. While these inconsistencies did not disrupt the system in 1987–91, the exchange rate stability of this period for the countries outside the core of the ERM is now recognized as an aberration, with markets (and governments) apparently falling spellbound in anticipation of EMU.

Second, the progress already achieved in nominal convergence has brought a shift of emphasis in what many countries need from the ERM. In particular, there is now a large core group of countries in the ERM that is characterized by broad nominal convergence. These countries are now less reliant on the ERM to support the disinflation process. As a result, the markets may see tensions between domestic policy goals and an exchange rate commitment as being more likely than in the past.

The third, and probably the key factor that sets the current environment apart from that of the 1980s, is the increased financial market integration that has resulted from the single market program and the abolition of capital controls. Many observers warned that the abolition of capital controls would, in the presence of underlying divergences in policies and performance, make the ERM fragile and potentially unstable: when markets take the view that the political will does not exist to sustain a parity, they can quickly marshal enormous sums to challenge it. During the recent crisis, the accumulated cost/price divergences and the tensions resulting from divergent relative cyclical positions presented the markets with one-way bets. Once they had made up their mind that a realignment was imminent, it was very difficult to convince them otherwise.

The recent experience has demonstrated how the sheer scale of funds that can be marshaled by the markets can cause problems for the defense of parities through interest rate measures. The size of the interest rate increases needed can be particularly problematic in the context of weak domestic conditions. The markets have become very aware of this. For example, doubts about whether the authorities would be able to raise interest rates significantly, in an environment where the desired path for interest rates was clearly downward, were important in the markets’ assessment of the willingness of French monetary authorities to fight off a speculative attack in the summer of 1993.

Against this background, the wisdom of seeking a return to the ERM as it existed up to September 1992 is widely questioned. Certainly, the ERM countries themselves are not rushing to reinstate such a system. They seem content, for the moment, with a situation in which currencies effectively fluctuate within narrow bands but without formal constraints. A more rigid system might re-focus the attention of markets and could be all the more fragile for having so recently been shown to be fragile.

While the system is formally much less constraining than that prior to August 1993, the countries that have remained in the ERM have, by and large, continued to emphasize the exchange rate orientation of policies. Many lessons of the recent crisis for country policies remain relevant—especially the problems of one-way bets that will continue to arise if countries pursue exchange rate policies that are not credible to the markets. Similarly, a number of issues related to the operation of the system are still germane.

Policy Requirements at the Country Level

Countries That Have Not Yet Converged with the Anchor. The ERM experience demonstrates that there are limits to the use of an overvalued exchange rate as a disinflation tool; eventually the financial markets will challenge such a policy. In a European context, in particular, it seems inevitable that the financial markets will now be less tolerant of countries that accumulate cost and price divergences than they were in the period leading up to the recent turbulence. Moreover, there may be significant risks for policy credibility in letting major disequilibria emerge. In particular, this is likely to elicit large-scale speculative pressures, and, by allowing the markets to appear in control, undermine confidence in policymakers and heighten uncertainty. Policymakers are likely to seem much more in control in a situation where a series of smaller corrections are smoothly implemented on a timetable of their own choosing.

The experience of the late 1980s and early 1990s underlines that a stable exchange rate on its own is not an effective tool for controlling inflation. This is particularly the case where structural rigidities provide shelter to the nontraded sectors of the economy over and above their natural insulation from outside competitive pressures. Indeed, at times, the problems of inflation may be compounded by the system itself. For example, a pick-up in inflation due to country-specific factors will often result in declining short-term real interest rates, as movements in short-term nominal interest rates are limited by the exchange rate peg, with the result that monetary policy acts procyclically.16 Thus, the exchange rate anchor needs to be supplemented by complementary actions in the fiscal and structural policy areas.

It is also clear that, in an environment of high capital mobility, it will be difficult to forestall speculative pressures when a devaluation is in prospect. In effect, the only way to discourage speculation is to convince speculators that there is no major profit opportunity. In a system where formal bands are an effective constraint on exchange rate movements, realignments ought to be frequent enough to avoid significant discrete movements in the exchange rate. Similarly, in a system with wider bands, policymakers will need to prevent the exchange rate from getting significantly out of line with underlying conditions. Moreover, policymakers need to be conscious of the focusing powers of significant dates on the political calendar and to avoid actions and statements that seem to call into question the policy strategy.

Countries That Have Already Converged with the Anchor. The policy needs for countries that have already achieved inflation convergence with the anchor country are in broad terms similar to those of the nonconvergent countries, i.e., they should foster flexible factor and product markets and ensure that fiscal policy is consistent with the exchange rate aims. The emphasis is somewhat different. While, in the nonconvergent countries, these instruments need to be marshaled to the fight against inflation, in the convergent countries, the weight is on ensuring that the economy is well placed to absorb real shocks without calling into question the commitment to the exchange rate anchor. In particular, the fiscal authorities will need to pursue prudent policies when economic conditions are strong in order to provide room for maneuver at times of weaker cyclical conditions. More generally, flexible factor and product markets will help absorb shocks.

As in the case of the nonconvergent countries, the role of fiscal and structural policies in supporting the exchange rate anchor can be reinforced by underlining the commitment to the rules of the game. Needless to say, policymakers should avoid statements that could in any way cast doubt on this. The granting of independence to a country’s central bank could strengthen further the credibility of exchange rate policies by reducing the potential influence of short-term political factors, a consideration that is of course also relevant for countries that have not yet converged.

Systemic Issues

A number of issues related to how the ERM operated prior to September 1992 have been the subject of debate. These issues continue to be important in a system of de facto narrow bands and are discussed below, taking first those that entail less radical proposals for changes to the system.

The Timing of Realignments and Spillover Effects. A central lesson of the ERM crisis is that management of a country’s exchange rate can have important spillover effects that make the timing of realignments a matter of common concern in a system such as the ERM. If a number of countries have let imbalances accumulate, there is a danger that generalized speculative pressures may call into question the system itself. Moreover, a substantial devaluation of an important currency can force spillover realignments in smaller trading partners that have relatively sound underlying conditions—realignments that might not be necessary where the country suffering the imbalances adjusts in a more gradual fashion. The spillover effects of the devaluations of the fall of 1992 on Ireland and Norway are typically presented as examples. These considerations reinforce the arguments, discussed above, for more frequent smaller realignments, rather than infrequent larger ones.

Realignments and Real Shocks. Through stabilizing nominal exchange rates, the ERM has also promoted an implicit form of policy coordination within Europe that can be helpful in responding to shocks that are experienced by all.17 However, the recent experience underlines the threat that asymmetric shocks can pose. From the perspective of the system, there are two aspects of such shocks that are worth stressing. First, as illustrated by German unification, a real shock hitting the anchor country has particularly wide-ranging consequences. Second, the rules of the game need to recognize that permanent real shocks can give rise to the need for a realignment, just as can delays in achieving cost/price convergence; traditionally, realignments in the ERM have focused on correction of accumulated cost and price differentials.

Considering exchange rate changes as a response to real shocks is seen by some as inconsistent with the spirit of the commitment to monetary union; after all, realignments will not be a policy option in a monetary union. However, if realignments are permissible to correct for delays in slowing inflationary momentum, it would also seem inappropriate to rule them out out as a response to a substantial and lasting real shock. Under both circumstances, of course, the nominal anchor will work most effectively if adequately supported by fiscal and structural policies, and realignments are resorted to only when the practical limits of other policies have been reached. Moreover, if asymmetric real shocks are a persistent problem, one might need to look at the appropriateness of an exchange rate anchor.

There are, however, important difficulties not just in recognizing the need for a realignment but also in calculating how large it should be and in organizing any needed complementary measures. The conceptual and practical problems in measuring real exchange rates are widely recognized.18 But the appropriate response to real shocks is an even more complex question. To illustrate this, it may be useful to look back at the issues that emerged in the context of German unification.

The benefits of a significant early general revaluation of the deutsche mark to absorb pressures raised by German unification have been well rehearsed. It would have reduced the initial demand boom in Germany, allowed the Bundesbank to ease interest rates at an earlier stage, and, overall, ensured a smoother path for output throughout Europe than has actually occurred.

Many of Germany’s ERM partners were, however, averse to a realignment. In significant part, this reflected the building enthusiasm for EMU. But there were also considerations of inflation and credibility. With output capacity in Western Europe already fully utilized, the increase in demand associated with German unification would entail higher inflation somewhere in Europe, in the absence of offsetting measures. A revaluation of the deutsche mark would in effect have shifted more of the increase in inflation from Germany to its partners. The increased demand would, of course, have been spread out over a larger economic mass, lessening its inflationary impact. Germany’s partners would, however, also have had to deal with the direct effect of devaluation through import prices. The other ERM countries’ partners were also concerned that a revaluation of the deutsche mark would undermine the future credibility of exchange rate policies and boost interest differentials vis-à-vis Germany.19 At the same time, they did not see the implications of increased aggregate demand in Germany as particularly damaging. They would face higher interest rates but this would be broadly offset by the benefit of stronger import demand in Germany.

It is clear that a cooperative strategy could have been devised to compensate for some of the unwanted implications of a general realignment. In particular, efforts could have been made in Germany’s European partners to offset much of the inflationary impact through fiscal measures; the increased demand for exports would have enabled governments to cut borrowing without incurring the cost of reduced output. The benefits that would have emerged later would have been derived not only from a less restrictive German monetary policy but also from increased flexibility in the fiscal area. Such a package, if clearly presented as a once-off set of measures to deal with an exceptional position, would not necessarily have entailed increased interest rate premiums against the deutsche mark. Moreover, in such a scenario, the size of the needed realignment would have been smaller if Germany had participated in such a fiscal effort by taking stronger measures to limit the increase in its own fiscal deficit.

The logistics of organizing a package such as that suggested in the preceding paragraph, however, would clearly have been quite complicated. Moreover, it is clear, with the benefit of hindsight, that any realignment package would have proved too small, given the widespread underestimation both of the unification shock and of the problems for economic policymakers in Germany that would subsequently emerge. The extent to which it would have alleviated tensions can only be a matter for conjecture.

Issues of Monetary Coordination. The recent ERM experience has also raised questions about the coordination of monetary policies in response to pressures in the system and, in particular, whether monetary policy in the anchor country should accommodate real shocks to some extent to minimize spillover effects on other countries. In a monetary union, of course, pressure in the monetary domain that emanates from one country is diffused by the union-wide orientation of monetary policy. It is clear that if such an orientation for monetary policy, targeted at low inflation in the ERM as a whole, had been in place in recent years, monetary conditions from the perspective of Europe as a whole would have been more appropriate. However, from Germany’s perspective, the inflationary consequences in Germany would have been undesirable. Moreover, as monetary policies remain the exclusive domain of national authorities, until monetary union is achieved, especially given that the German monetary authorities are legally obliged to pursue price stability in Germany as their principal goal, it would have been unreasonable to expect the Bundesbank not to put heavy emphasis on inflation in Germany.

Given the close integration of European economies, the anchor country in pursuing its own goal of price stability does in effect take into account economic conditions in other countries, but only to the extent that they affect its own inflation rate. Its monetary authorities, with their strong anti-inflation credentials, can no doubt from time to time give additional weight to circumstances in neighboring countries, without affecting their credibility or the strength of the anchor. However, once the anchor country begins to adapt its policies to the needs of follower countries on a regular basis, or for a sustained period, especially in face of domestic inflationary pressures, its role as anchor begins to become blurred. If there is a consensus that a coordinated approach to policy is to become a general feature of policies, a common anchor would be required, such as the common monetary policy that will underpin EMU. Other papers in this book deal with monetary union in Europe.

The Choice of Nominal Anchor. The exchange rate anchor has lost some of its lustre after the recent year-long crisis in the ERM, rekindling the debate on the appropriate orientation for monetary policy in European countries and with this debate have come proposals for alternative anchors. While a detailed consideration of such issues goes beyond the scope of this paper, a few comments may be helpful. First, as discussed earlier, the nominal anchor is not the only reason for seeking exchange rate stability. Second, the experience of Austria and the Netherlands underlines the fact that nominal exchange rate pegs are not necessarily fragile. In both countries, the strength of the peg is assisted by a very close integration with the German economy; in recent years, for example, the inflation pattern in both countries has followed that of Germany more closely than was the case for other countries. However, the long history of following the deutsche mark and the clear disinclination of policymakers to cast doubt in any way on the link have also been critical.

Third, alternative anchors also suffer from drawbacks. The money supply has long been seen as the principal alternative to the exchange rate as a nominal anchor. For many countries in Europe, however, volatility in the velocity of money represents a strong argument against such an approach. Some European countries with strong records of relative price stability—in particular, Germany and Switzerland—have relied on monetary targeting to provide the general orientation to policy. These, however, are seen by many as being special cases, where the long-standing credibility of the monetary authorities has allowed them significant discretion in the implementation of policy. The changes in financial markets related to the ongoing process of market integration in Europe are seen to weaken further the case for monetary targeting, at least at the level of the individual country. It does, however, remain a candidate for the operating framework to be chosen by a future European central bank.

An alternative approach that has been gathering support in recent years is the direct targeting of inflation. The principal drawback to this is the long lag between monetary policy action and its reflection in price behavior, implying significant discretion for policymakers and less transparency in policy. It is argued by proponents that credible inflation forecasts, combined with clear guidelines as to how monetary policy should respond when future inflation projections deviate significantly from target, can offset these weaknesses. At the same time, they might allow policymakers to avoid some of the problems that can be associated with exchange rate anchors, and in particular conflicts between the needs of the domestic economy and external constraints. In Canada and New Zealand, the successful reduction of inflation has been assisted by attempts to guide price expectations through the announcement of formal inflation targets and in the United Kingdom, such a policy has been adopted following the withdrawal of the pound sterling from the ERM. The long-term effectiveness of this approach in guiding inflationary expectations will, no doubt, depend on establishing a successful track record. It is still too early to determine the specific contribution of explicit targets to the recent reduction of inflation in Canada and New Zealand.

Concluding Remarks

Following the difficulties of 1992–93, the ERM clearly has lost some of its appeal. Its influence has diminished somewhat, with two large countries having withdrawn, and three Nordic nonmember countries having delinked their currencies from the ECU. Moreover, the vulnerability of fixed exchange rate parities in an environment of mobile capital is now widely accepted. With many countries already having achieved low inflation, the ERM’s role as a disinflation tool has also receded in prominence.

Despite the difficult experiences of 1992 and 1993, the frequent talk of “fault lines,” and the widening of the bands, the ERM has reemerged with currencies effectively adopting narrow fluctuation margins against the deutsche mark. The ERM’s durability is, in part, testimony to the broader importance attached to stable exchange rates in Europe. This has assumed even greater weight in the transition to EMU. Not only are stable exchange rates during the last two years of the transition to EMU required under the Treaty on European Union, but they are also seen by many in Europe as a way to demonstrate that a monetary union is sustainable. Moreover, with possibly only a matter of a few years until EMU would be realized, there has been great reluctance, especially among the core ERM countries, to give up the large investment in the existing approach to monetary policy and shift to an alternative whose credibility might take some time to establish.

The prospective shift to EMU has raised important questions as to the nominal anchor to be used in the future, with the exchange rate unlikely to be an attractive option for the union as a whole. While monetary targeting will no doubt be a leading contender, the ERM turmoil has also focused attention on inflation targeting as a candidate. By the time a decision on EMU is made and the future European Central Bank selects its operating procedures, there is likely to be somewhat more information on how a number of countries have been able to make use of this alternative approach.


    Artis, Michael J., and MervynK. Lewis, “Après Le Déluge: Monetary and Exchange-Rate Policy in Britain and Europe,”Oxford Review of Economic Policy,Vol. 9, No. 3 (Autumn1993), pp. 36–61.

    • Search Google Scholar
    • Export Citation

    Calvo, Guillermo, and CarlosA. Végh, “Inflation Stabilization and Nominal Anchors,”IMF Paper on Policy Analysis and Assessment, No. 92/4 (Washington: International Monetary Fund, December1992).

    • Search Google Scholar
    • Export Citation

    deGrauwe, Paul, “The Cost of Disinflation in the European Monetary System,”CEPR Discussion Paper Series, No. 326 (London: Center for Economic Policy Research, 1989).

    • Search Google Scholar
    • Export Citation

    deGrauwe, Paul, The Economics of Monetary Integration (Oxford and New York: Oxford University Press, 1992).

    Egebo, T., and A.S.Englander, “Institutional Commitments and Policy Credibility: A Critical Survey and Empirical Evidence from the ERM,”OECD Economic Studies,Vol. 18 (Spring1992), pp. 45–84.

    • Search Google Scholar
    • Export Citation

    Giavazzi, Francesco, and AlbertoGiovannini, Limiting Exchange Rate Flexibility: The European Monetary System (Cambridge, Massachusetts: MIT Press, 1989).

    • Search Google Scholar
    • Export Citation

    Giavazzi, Francesco, and AlbertoGiovannini, and L.Spaventa, “The New EMS,” in The European Monetary System in the 1990s,ed. by Paulde Grauwe and L.Papademos (New York: Longman, 1990).

    • Search Google Scholar
    • Export Citation

    Goodhart, Charles, “Economists’ Perspectives on the EMS,”Journal of Monetary Economics,Vol. 26 (December1990), pp. 471–87.

    Gros, Daniel, and NielsThygesen, European Monetary Integration: From the European Monetary System to the European Monetary Union (London: Longman, 1992).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund, International Capital Markets: Part I. Exchange Rate Management and International Capital Flows (Washington: International Monetary Fund, 1993).

    • Search Google Scholar
    • Export Citation

    Kiguel, MiguelA., and NissanLiviatan, “The Business Cycle Associated with Exchange Rate-Based Stabilizations,”World Bank Economic Review,Vol. 6 (May1992), pp. 279–305.

    • Search Google Scholar
    • Export Citation

    Lipschitz, Leslie, and DonoghMcDonald, “Real Exchange Rates and Competitiveness: A Clarification of Concepts, and Some Measurements for Europe,”IMF Working Paper, No. 91/25 (Washington: International Monetary Fund, 1991).

    • Search Google Scholar
    • Export Citation

    Sachs, Jeffrey, and CharlesWyplosz, “The Economic Consequences of President Mitterrand,”Economic Policy: A European forum,Vol. 1, No. 2 (April1986), pp. 261–322.

    • Search Google Scholar
    • Export Citation

    Salop, Joanne, “The Divergence Indicator: A Technical Note,”Staff Papers, International Monetary Fund, Vol. 28 (December1981), pp. 682–97.

    • Search Google Scholar
    • Export Citation

    Spaventa, L., “The New EMS: Symmetry Without Coordination?” in The EMS in Transition (London: Center for Economic Policy Research, 1989).

    • Search Google Scholar
    • Export Citation

    Tsoukalis, Loukas,The New European Economy: The Politics and Economics of Integration (New York: Oxford University Press, 1993).

    Ungerer, Horst, and others, The European Monetary System: Recent Developments, Occasional Paper, No. 48 (Washington: International Monetary Fund, 1986).

    • Search Google Scholar
    • Export Citation

    Ungerer, Horst, and others, The European Monetary System: Developments and Perspectives, Occasional Paper, No. 73 (Washington: International Monetary Fund, 1990).

    • Search Google Scholar
    • Export Citation

    Walters, AlanA.,Britain’s Economic Renaissance: Margaret Thatcher’s Reforms, 1979-1984 (New York: Oxford University Press, 1986).


The pound sterling, and with it the Irish pound, was a member for only two months, the lira for less than a year, and France’s intermittent membership totaled less than two years.


The ECU comprises a weighted basket of different amounts of the EMS currencies. The basket was periodically redefined during the 1980s to take account of realignments and new EMS members. Under Article 109g of the Treaty on European Union (the Maastricht Treaty), the currency composition of the ECU is fixed during Stage II of monetary union, which began in January 1994.


Although all EC members joined the EMS, they were not required to join the ERM. The original ERM members were Belgium, Denmark, France, Germany, Ireland, Italy, Luxembourg, and the Netherlands. The Italian lira was allowed a 6 percent band, which it followed until it shifted to the 2.25 percent band in January 1990.


Such transactions were to be recorded at the European Monetary Cooperation Fund (EMCF); after two and a half months, repayment would become due, half of which would be paid in the currency of the creditor; the other half could be repaid in ECUs. The initial loan could be of an unlimited amount; renewal for a further three months was limited to the amount of the debtor country’s quota in the short-term monetary support facility. The functions of the EMCF have now been taken over by the European Monetary Institute (EMI).


There was nothing to prevent central banks of countries with strong currencies from sterilizing intervention at the margin and the divergence threshold proved a weak instrument, principally because it created only a presumption of action, not a requirement, but also because a country with a large ECU weight was less likely to cross this threshold. For further details consult Giavazzi and Giovannini (1989), de Grauwe (1992), Salop (1981), Tsoukalis (1993), and Ungerer and others (1986, 1990).


For further discussion see Goodhart (1990) and Gros and Thygesen (1992).


Sachs and Wyplosz (1986) and Ungerer and others (1986).


The Basle-Nyborg agreement of September 1987 extended the duration of VSTF loans by one month to three and a half months, doubled the quota that received automatic renewal, increased the acceptance limit for ECUs in VSTF settlements to 100 percent, and extended the scope of the VSTF, subject to limits, to include intra-marginal interventions. It also provided for increased surveillance of the economic policies of member countries and emphasized the use of interest rates to defend parities. It was agreed that exchange rate realignments should be infrequent and as small as possible. See Ungerer and others (1990) and Giavazzi and Giovannini (1989).


There was one technical realignment when the central rate of the Italian lira was devalued, without changing the lira’s lower intervention limit, when its intervention margins were narrowed to 2.25 percent in January 1990.


A fifth nonmember, Austria, had effectively pegged its currency to the deutsche mark since 1981.


de Grauwe also found that the “misery index,” defined as the sum of the unemployment rate and the inflation rate, worsened more for ERM countries than for non-ERM countries.


This distinction is similar to that made between exchange rate and money-based stabilization in high-inflation countries. See Kiguel and Liviatan (1992) and Calvo and Végh (1992).


This is the conclusion reached by Goodhart (1990). Artis and Lewis (1993), and Spaventa (1989), among others.


“See International Monetary Fund, World Economic Outlook, May 1993 and October 1993, the Interim Assessment of the World Economic Outlook, January 1993, and International Capital Markets: Part I. Exchange Rate Management and Capital Flows, April 1993.


The German current account position shifted by 6 percentage points of gross domestic product (GDP) between 1989 and 1991.


This is known as the Walters critique (see Walters, 1986). The recent problems for the low-inflation ERM countries posed by “high” German rates are analogous.


As an illustration, suppose that an inflationary demand shock hits the European countries. If restrictive monetary policies are introduced by each country independently, the result may be an unpredictable and volatile pattern of exchange rates. Pegging the exchange rate facilitates a coordinated response.


See, for example, Lipschitz and McDonald (1991).


There had been no previous case of a realignment effected on the basis of ex ante considerations; all previous realignments had been intended to correct competitiveness losses that had already occurred.

    Other Resources Citing This Publication