Answering this question supposes an assessment of the EMS, as it worked between 1979 and 1993 - that is when the EMS ceased to exist for all practical purposes - even though it continued to exist legally - since the fluctuation margins where widened up to +/- 15 percents.

I The credibility problem (first structural flaw)

Pegged exchange rate systems face some important problems that have led many economists to doubts the long run sustainability if these system :

- - A first problem as to do with credibility of the fixed exchange rates

- - A second has to do with the way the system-wide monetary policy is determined. It is therefor a problem of the monetary policy making process (which leads to tensions)

A - Why credibility is important ?

1 - Credibility influences exchange rate expectations...

The exchange rate expectations of economic agents are shaped by the level of trust they put into the public authorities declared behaviour.

A policy is credible if there is no better alternative available to the public authorities. If it is not the case, it is likely that the public authorities will change their behaviour soon. Therefore, their policy is not credible.

2 - ... and exchange rate expectation are self fulfilling

When it comes to exchange rates, expectations of economic agents are self-fulfilling. This is why credibility is an important issue for central banks. We shall argue that there are two sources of lack of credibility : the short term incentives for national government to renege their commitment, and the different level of reputation that a given central bank has attained.

B -The sources of low credibility

1 - Incentive for national government to renege their commitment

The credibility problem appears in situation whereby the cost of relinquishing the exchange rate instrument seems to outweigh its benefits. It arises for two different reasons. Let us assume that a wage explosion occurs in France (like in 1968). This entails an upward movement of the supply curve. It leads to a current account deficit and a reduction of the French output and employment. If wage flexibility and labour mobility are low, and given the fact that France is committed to no devaluation, this country faces a dilemma.

It can reduce the current account deficit by following a deflationary monetary and fiscal policies (at the cost of lower output and employment). Or it can increase output and employment by expanding monetary and fiscal policies (at the cost of a higher current account deficit).

It is clear that the incentive to use the exchange rate instrument to solve the dilemma is very high. (Indeed, a devaluation could, in a Keynesian view, increase the French competitiveness and therefore boost the output, employment and then reduce the deficit). The authorities have an incentive to renege on their commitment to keep the exchange rate fixed.

Speculators will realise that the government has this incentive. As a result they will believe it is likely that a devaluation will occur, and therefore could make it actually occur (expectations of economic agents are self-fulfilling).

This credibility problem can only be solved if the authorities convince speculators that their only objective is the fixity of the exchange rate (whatever the cost in terms of output and unemployment). It is hard to make such a commitment.

2 - Difference in reputation

The BARRON - GORDON model stresses that the degree of credibility of national monetary authorities is essential to ensure a low inflation rate. The expectations of economic agents are, indeed, self-fulfilling.

Therefore, if a national monetary authority, say Italy, are perceived as 'wet' and short-sighted (and not 'hard-nosed'), economic agents will anticipate devaluation. This leads to large scale speculative crisis prior to the anticipated date of re-alignment.

Different levels of credibility among participating central banks (Germany, Italy) can therefore lead to structural and permanent desequilibria.

II The liquidity problem (second structural flaw )

A - Monetary control in a pledged exchange rate system : 2 possibilities

1 - The n - 1 problem : existence of a degree of liberty

Every system of fixed exchange rates faces the problem of how to set the system-wide level of the money stock and interest rates. This issue arises from the so-called n - 1 problem. In a system of n countries, there are only n - 1 exchange rates. Therefore, n - 1 monetary authorities will be forced to adjust their monetary policies so as to maintain a fixed exchange rate. There will be one monetary authority which is free to set its monetary policy independently.

Þ the system has one degree of freedom. What central bank will use it ?

It is important to realise that the fixed exchange rate arrangement is compatible with any possible level of the interest rates and of money stocks. There is a fundamental indeterminacy in the system.

2 - The asymmetric (hegemonic) solution

One solution consists in allowing one country to take the leadership role. The country A is the leader and fixes its money stock independently. This then fixes the interest rate in country A at the level R1. Country B now has no choice any more. Its interest rate will have to be the same. Given the money demand, this uniquely determines the money supply in country B.

Þ country B has to accept this money supply. It cannot follow an independent monetary policy.

Þ Country A takes the anchoring role.

3 - The symmetric(co-operative) solution : original blueprint of the EMS

A second possibility is for the two countries to jointly decide about the level of their money stocks and interest rates. Thus, this solution requires co-operation.

B - EMS : asymmetric monetary control (no co-operation)

1 - The EMS was meant to promote the co-operative solution

The original blueprint of the EMS was aimed at promoting this co-operative solution. Two features were intended to promote that :

1. The ECU, as a basket currency, was a symbol of political equality

2. The symmetric intervention system : when two currencies hit the upper limits of their cross parity, intervention should be undertaken by both central banks

3. automatic financial co-operation : central banks must co-operate and offer indefinite credit lines to central banks which need it.

This would require a country with a strong currency to expand its monetary policy, and a country with a weak currency to contract its monetary policy.

2 - However, symmetry did not come about : Germany has been unwilling to pay the price of intervention (expansion of its money supply)

In particular, when a speculative crisis arose, requiring intervention in the foreign exchange market, the strong currency country (Germany) has generally been unwilling to allow its money stock to increase and its interest rate to decline.

3 - Sterilisation policies Þ the weak currency is forced to do all the adjustment

This Germany has achieved by using sterilisation policies. They consist in offsetting the expansionary effects of the interventions on the foreign exchange market by reverse operations of the Bundesbank. Thus, when a central bank of the weak currency country sold marks (against its own currency), the Bundesbank has usually bought these marks back through open market operations.

The effect of these sterilisation policies was that the German money stock was not affected by the operations of the weak-currency countries.

The implication of this asymmetry is that the weak currency is forced to do all the monetary adjustment.

Conclusion : despite the original intention of the founding fathers of the EMS, the system has evolved in to an asymmetric one whereby Germany sets up the system wide monetary policy.

C - The problems of no co-operation : the liquidity problem

1 - Asymmetric shocks in an asymmetric monetary control

a) The open interest parity : the interest rate must be the same everywhere in the system (due to capital mobility)

This is because of the 'open interest parity' due to the assumed perfect mobility of capital. This condition says that the interest rates between two countries A and B. The condition is :

Ra = Rb + m

This says that if the economic agents expect a depreciation of currency A, the interest rate of country A will have to exceed the interest rate of country B by the rate of the expected depreciation. If agents do not expect any adjustment, there is no expected depreciation between countries. Therefore m = 0 and Ra = Rb.

b) recession in peripheral countries are amplified

Unsynchronised business cycles can have grave consequences in an asymmetric pegged exchange rate system, whereby one anchor country follows its own monetary policy without co-operating with the peripheral countries.

A recession which originates in the peripheral countries is made worse by a contraction of the country's money stock. In order to show that, let us assume that a recession occurs in a peripheral country. Consequently, the money demand shits downwards ( Md = P.L(Y,r) ). This has the effect of reducing the interest rate.

Since the anchor country experiences no change in its money demand, its authorities do not feel the need to change the money supply, which means that its interest rate remain constant.

The result is that the money supply in the peripheral country must automatically decline This comes about as follows. The downward pressure on the peripheral country's interest rate leads to an outflow of capital to the centre country. This reduces the money stock.

The non adjustment of the interest rate and the deflationary effect on the money supply make worse the initial recession in the peripheral countries.

c) expansions in the peripheral countries are amplified

The opposite occurs when a boom originates in the peripheral countries. In that case, the boom is automatically accommodated by an increase in the money stock of the peripheral country, because the upward pressure on the interest rates leads to capital inflow.

2 - Asymmetric shocks in an symmetric (co-operative) monetary control

The symmetric system of monetary control would be more successful in stabilising the system's money stock when unsynchronised business cycles occur.

We now assume that the central banks of the centre and the periphery co-operate to stabilise the whole system's money stock, in case of a recession in the peripheral country. The peripheral country reduces its money stock and the centre country increases its money stock. The total money stock keeps unchanged but the interest rate decreases.

2 - Given these economic effects, asymmetric monetary control leads to conflicts

The problem of monetary control that arise in an asymmetric system when asymmetric shock occur are likely to lead to conflict about the kind of monetary policy to be followed for the whole system.

Peripheral counties, especially if they are similar in size to the centre country may not be willing to subject their national interests to the survival of the system.

Þ in the end, more explicit co-operation. The move towards monetary union can be interpreted in this sense.

II What makes a pegged exchange rate system last ? The example of the EMS

The surprising about the EMS is that it lasted for more than a decade before it succumbed to the problems we have just identified. What were the factors that helped the system temporarily to overcome the problems discussed. Two features are important here.

A - Bands of fluctuation

First there is the existence of the band of fluctuation. Contrary to the Bretton Woods system they where were relatively narrow (1 percent above and below the parities), in the EMS the bands were fixed at 2.25 percents. For some countries the bands were event 6 percents (For Italy until July 1990, Portugal, Spain, and the UK).

These relatively large bands of fluctuation were quite important in stabilising the system. Indeed they make possible for the monetary authorities of high inflation countries to change the exchange rate regularly by small amounts without having to face large speculative crises (prior the expected realignment).

It is important to note that the bigger the jumps realignments entail, the bigger the profit opportunities for speculator, the more likely are important speculation crisis. The large bands of the EMS (as compared to Bretton Woods) reduced the occurrence of speculation crisis.

The relatively small and frequent realignments helped to reduce the size of speculative movements. This feature of the EMS, enabled monetary authorities to retain some monetary flexibility, but change dramatically after the Basle-Nyborg Agreement of 1987

B - Capital control

The early EMS was also characterised by the existence of capital control. In particular in France and in Italy, it tended to reduce the size of funds mobilised for attacking currencies. They basically gave authorities time to organise orderly realignments.

At the end of the 80s, capital control was being progressively abolished, as a step towards a complete market integration. This move changed the nature of the EMS.

III The disintegration of the EMS

At the start of the 1990s, the EMS has evolved into a truly exchange rate system with almost perfect capital mobility.

The first consequence, that can be understood using the incompatible monetary triangle (Mundell), was that monetary autonomy was completely abolished for all the countries, apart for the anchor country (Germany). Indeed capital (mobility + fixed exchange rates) result in no monetary autonomy.

In this new EMS the problems of credibility and of liquidity started to have their full destabilising effects.

A - The lira and the peseta crisis : a credibility problem

A first problem arose with the lira and the peseta. Despite vigorous and relatively successful efforts at reducing the inflation in Italy and Spain., these two countries did not manage to close the gap with Germany. The credibility problem is the cause and the consequence of this fact.

1 - Lack of credibility of central banks

Indeed, this situation has certainly something to do with differences in the reputation of the monetary authorities in Italy and Spain, which made it difficult to reduce inflationary expectations.

2 - An ever increasing loss of competitiveness

This situation leads to a credibility problem in the second sense. As the Italian and Spanish inflation rates remained above the German one for many years, the price levels continuously diverge. Since there were no realignment after 1987 to compensate for these divergent prices, a continuous loss of competitiveness of the Italian and the Spanish occurred.

Þ This loss of competitiveness (of the Italian and the Spanish industries) reached up to 30 percents. In the end this became unsustainable.

B - The sterling and the French franc crisis : a liquidity problem

1 - No credibility problem : the fundamentals were good

The problems that arose with the French franc and the pound sterling were different in nature. Unlike for Italy and Spain, the 'fundamental' economic variable of France and the UK were comparable to those of Germany. There was no structural or permanent loss of competitiveness.

2 - The 'Anglo Saxon Plot'

Some continental European observers and politicians have claimed that the speculation against the pound sterling and especially against the French franc was irrationally driven by an 'Anglo Saxon plot' against the process of monetary unification in Europe.

3 - A liquidity problem triggered by the 1990s recession : a policy conflict

A better explanation is available, base on the liquidity problem of fixed exchange rate regimes.

At the start of the 1990s and especially after 1992, Europe was hit by a sever recession. Very quickly, this created a conflict, between Germany on the one hand, and Britain and France on the other hand, about the appropriate interest rate policy to be followed in the system as a whole.

Things were complicated by the German reunification of 1990 which had led to increases in government spending in Germany which created inflationary pressure. The latter were reinforced by the one-to-one exchange rate policy between eastern and western DM.

As a result, the Bundesbank gave complete priority to combating inflation by a restrictive monetary policy. The recession in the UK, France, however, demanded a loose monetary policy Þ conflict.

4 - Speculators noticed the policy conflict

This policy conflict did not remain unnoticed by the speculators. These realised that the UK and French authorities were tempted to cut their links with the mark so as to be able to follow more expansionary policies. Influential economists openly urged the government to do so. Speculators had good reasons to star speculating against the pound and the sterling

5 - The crisis

This happened first in September 1992 and led to the withdrawal of the pound from the ERM of the EMS. In 1993, a new speculative crisis erupted involving the French franc (but also the peseta, the Belgian franc and the Danish kroner).

The second of August, the margins of fluctuation were increased to +/- 15 percents, transforming the EMS into a quasi floating exchange regime. Although in a legal sense the EMS remained in existence, for all practical purposes the system ceased to exist.

C - Is speculation responsible ?

1 - The cases for their responsibility

1. Indeed, objectively speaking, speculators forced the authorities to drop out of the system.

2. speculation is fulfilling in nature. That is to say, expectations tend to become true by themselves.

3. the 'Anglo Saxon' plot against monetary unification in Europe.

This features led many people to think that speculators are to blame for the collapse of the EMS.

2 - The case for their non responsibility :

The EMS was fundamentally unstable, due to the credibility and liquidity problem of fixed exchange rate systems.
it is doubtful that capital controls would have prevented the disintegration of the EMS. After, all the Bretton Woods system collapsed for essentially the same reason, despite the fact that capital controls existed at the time of its collapse.
It is not true to say that speculators can mobilise a vaster amount of liquid funds than the central banks can. This view is wrong. Against the large amount of money there is an equally large amount of money the authorities can throw in the battle. Central banks can create an unlimited amount of money. If they co-operate, the monetary authorities can always win the battle against the speculators. This is were Germany played a key role.

D - The role of Germany

Why did the monetary authorities lose then ? The answer is that they did not co-operate fully. German authorities refuse to continue supplying the French authorities with German marks. This refusal came about because Germany was unwilling to significantly increase its money supply (German money stock). Indeed the size of the intervention was so high that Germany was not capable of completely sterilising them. Thus, a massive support of the French franc would have forced the Bundesbank to relax its monetary policy stance.

CONCLUSION

The fundamental reason for the disintegration of the EMS can be said to be the conflict about the appropriate monetary policy in the system and the refusal of Germany to follow a more expansionary policy.

Our conclusion is therefore that a fixed exchange rate mechanism has profound and structural flaws that makes it bound to be a mere temporary agreement - as it happened to be for the Bretton Woods system and more recently for the EMS.

BACK TO PUBLIC POLICY ANALYSIS