The logic of supra nationality:
How many stages in Stage III ?

Marc Flandreau (*)

It is well known that the Maastricht Treaty has envisioned a process of monetary unification whose final stage (Stage III) will "irrevocably fix the exchange rate at which the ECU would replace the national monies" (Art. 109 L). The recent Maas Report (1) on the other hand argues that this stage will be itself decomposed in two steps because "although exchange rates will be irrevocably fixed on the first day, national currencies will keep their function for a while before the single currency can be introduced physically" (p. 5). At first sight, this suggests that the full completion of the monetary union would have to wait until the end of Stage III. This paper is an attempt to clarify this apparent discrepancy by discussing the logic of supra nationality in the making of a stable exchange rate agreement.

There are two sorts of stable exchange rates agreements. One is based on centralization: such regimes are operated through a set of rules going "top-down", from the center to the regions (or nations). The other type is based on national sovereignty, and relies on horizontal controls from regions to regions. The transition from the current EMS system to the implementation of stage III as envisioned by the Maastricht Treaty corresponds to the attempt at moving from the latter system to the former one. This will imply a change from horizontal to vertical rules: in other words, the contractual basis of the current EMS is about to be deeply modified.

In order to be able to locate the turning point of this endeavour, and in particular to clarify what is meant by saying that national currencies will keep their functions for a while, we provide a discussion of the architecture of the current EMS system and of the forthcoming regime that will emerge from the beginning of Stage III. Our claim is that the ECU is about to be transformed from what may be viewed as a mere appendix of the current regime, to the a fully fledged currency which will replace national currencies, not at the end of the unification process, but on day one of Stage III, even if national units remain transitorily in use. Besides, we argue that at this point, the ECU will become the main reserve asset of European nations which will not be able to join the process of monetary unification at its earliest stage: this situation, which was not fully envisioned by the Maastricht Treaty - the Treaty implicitly assumed that a large number of countries would be able to join - deserves close scrutiny, as the prospect of a "two speed Europe" becomes more likely. In both cases, the ECU will emerge as the monetary standard of the European Community.


Decentralized arrangements for exchange rate fixity or quasi-fixity such as the EMS necessarily combine three elements (Fitoussi & Flandreau 1994).

The most obvious one is the "principle of stability", which may be materialized by defining bilateral parities, setting target exchange rates, or placing limits upon exchange rate fluctuations.

The second key principle may be dubbed "principle of sovereignty": it means that countries remain sovereign in the definition of their monetary and budgetary policies. Besides, the principle of sovereignty indicates that nations are free to opt out of the arrangement if they decide that their monetary policies are no longer compatible with those followed by other members.

The third principle is the "principle of guarantee" which says that limits have to be put upon the amount of another country's currency that each nation has to absorb.

Intuitively, the third principle may be understood as an implication of the first two ones. Assume for instance that a member decides to adopt a target rate of inflation which is well above the one prevailing in other countries. According to the principle of stability, other members will be forced to intervene in support of the "inflationary" member, thus becoming themselves inflationary (2). On the other hand, the principle of sovereignty implies that countries can decide to opt out, i.e. refuse to provide unlimited support to "weak" currencies.

The principle of sovereignty and the principle of stability open an acute dilemma between external and internal stability (i.e. between inflation and exchange rate fixity). This dilemma in turn requires that a third principle be added to the first two ones: by allowing every participant to place upper limits upon the backing that one is willing to provide to foreign currencies, the principle of guarantee renders participation to a decentralized exchange rate arrangement acceptable (3). Hence, the role of the principle of guarantee is to rule out inflationary policies, by rendering bilateral exchange rate stability conditional upon the adoption of monetary policies which remain in line with those of the more conservative members.

From an economic perspective, the principle of guarantee is an incentive mechanism which forces every participant to internalize the externalities of its monetary policy upon the other member states.

The main shortcoming of this type of system however, is that regardless of the symmetrical nature of the rules which govern it (every actor faces the same rules of the game), it is bound to operate asymmetrically. This arises because countries have different tastes about the rate of inflation that they prefer, and derive different gains from participating to the currency union. For instance, countries with a reputation for preferring a high rate of inflation will normally face a credibility problem which will be eliminated by participating to a currency union: this implies that their gain from being in this arrangement will be larger than that obtained by countries with no credibility problem.

As a result, decentralized monetary union's will tend to converge towards an inflation rate which will be closer to that preferred by members who gain less from participating to the arrangement. Decentralized unions tend to produce a deflationary bias.


As mentioned in the introduction, the features characterizing the first step of the transition to Stage III, as envisioned by the Maas report, suggest that one of the most visible aspects of the decentralized regime which prevails today will still be in force: households, companies and governments will continue to conduct transactions in national currencies. We believe however, that this superficial fact hides a deep transformation of the whole architecture of the system, which will in practice move to a de facto monetary union on day one of Stage III: from that point on, the creation of a supranational monetary institution will be materialized by the transformation of the ECU from a basket into a fully fledged currency.

To see this, it is important to recall what will happen at the very beginning of the last stage. Indeed, from that point on, the banking and financial institutions will perform the role of transforming ECUs into national units, and vice versa. They will become a two way street operating at a fixed exchange rate between national currencies and ECUs. However this full "convertibility" - or perhaps more accurately full equivalence - between ECUs and national units will not at first take place through physical exchange of, say DM for ECU. Rather, national units and ECU will exist in different compartments of the payment system, and they will be merely re labelled when moving from one compartment to the other one: for instance, money will be denominated in e.g. FF. from the point of view of French households, but in ECU from the point of view of French banks. Credits granted in ECUs, will be received in francs, and will produce deposits in francs that will automatically become ECUs.

Of course, the smooth operation of such a system imposes a number of rules, because it grants national institutions an unlimited power to convert national units into the union's currency, which the other banks have to accept. In other words, this regime has no principle of guarantee. Suppose for instance that the Bank of France is allowed to hold unlimited amounts of French debt. In this case the Bank of France can create francs on the basis of French government's debt. But these francs are being automatically transformed into ECU: monetary creation in France gives rise to a corresponding ECU creation which would escape from "European" control. And, if the Bank of France is ready to hold unlimited amounts of French debt, it can as well create unlimited amounts of ECUs.

There is only one way to solve this dilemma, while maintaining automatic transformation of national units into ECUs and vice versa. It is to place upper limits upon the amount of public debt that central banks can purchase on the primary market (4): this is what Article 104 of the Maastricht Treaty achieves, although through a channel that is probably overly restrictive, since the upper limits it places upon direct purchases of public debt is set equal to zero. In other words, the removal of the principle of guarantee, which will happen on day one of Stage III, is complemented by the removal (or rather, strong limitation) of the principle of sovereignty. National central banks will no longer be autonomous in the formulation and implementation of monetary policy: a monetary union will be born.

This in turn shows that the making of supra nationality is not materialized by the removal of national monies as units of account, or even as physical means of payments in the shape of national bank notes. Instead, it is materialized by a set of rules, which transfer the monetary power from the regions to the center. DM, FF., etc. may nominally exist for a while after the beginning of Stage III. They may even remain in use for an extensive period of time. They will nevertheless be mere names - not monies - from the very first day of Stage III.


The implementation of Stage III will not only be important for countries which will be able to join in at the beginning of the process. It will also be important for outsiders who will have to wait until they achieve a greater level of convergence. Indeed, transforming the status of the ECU from a simple basket to an effective currency (i.e. a claim on the economies of participating countries) will dramatically modify the situation of outsiders because their reserves in currencies of insider countries will automatically become indirect ECU claims. This will put the ECU at the center stage of the European Monetary System, for both insiders and outsiders. Intervention on foreign exchange markets as well as exchange rate management, instead of incorporating the multilateral structure of the current system, will be explicitly dominated by bilateral relations between the European monetary union "core" (characterized by vertical rules and an effective transfer of sovereignty to supranational institutions), and the union's "periphery" (made of countries which will peg their exchange rates in terms of ECUs).

Here again, of course, the three principle of stable exchange rates arrangement will prevail: exchange rate stability, sovereignty of non-ECU countries vis à vis ECU countries, and contractual relations regarding the amount of support that ECU countries will provide to the periphery. This of course, raises the question of possible asymmetries in the effective operation of this regime. Given the disproportion that will exist between the core and the periphery, such difficulties are at least likely, and will probably arise as a by product of the "bad boys stay out" stance of the Maastricht Treaty.

On the other hand, it should be emphasized that the Maastricht Treaty had recommended that until the beginning of Stage III, "each country should treat its exchange rate policy as if it were a problem of common interest" (Art. 109 M). Of course, given the prevalence of the principles of guarantee and sovereignty until the end of Stage II, it has been, it is, and it will be difficult to enforce such recommendations. After the beginning of Stage III however, the situation is liable to be quite substantially modified. Once the process of unification will be on for some countries, the likelihood of more countries joining in will increase. And insiders will probably realize that it may be wise to treat the policy of the union vis à vis outsiders as a problem of common interest. To what extent should this issue be discussed in advance is a question that deserves some attention, especially because the number of outsiders might be larger than what the Maastricht Treaty had initially assumed.


There are several implications to be drawn from the previous analysis. In particular, despite the apparent gradualism which transpires from the Maas report, it should be emphasized that the main step in the completion of the European monetary union will be completed on day one of Stage III, not when national currencies are replaced everywhere by ECUs.

This in turn suggests that the accompanying measures (regarding for instance the legal tender status of national currencies in other countries) are more technicalities than essential questions. In particular, given that the various national monetary institutions will operate under the authority of the ECB, and that the ECSB will use ECU claims as clearing instruments, there is no need to introduce additional legal tender legislation. Intra-ECU zone transfers will take place in ECU through the agency of the European banking system which will operate as a transforming machine from national units to national units through the agency of ECUs. As for the convenience of travellers, it could be provided for by obliging national central banks to exchange at par foreign bank notes for national ones.

Finally, we have argued that Stage III will to some extent be a transitional phase, but only for countries that will not be able to join in at the beginning of the process. In this case, gradualism will prevail in the evolution of their bilateral links with the ECU area. Clarifying the future relations between the European union "core" and its "periphery" is an issue which should retain the attention of both policy makers and economists.



(*) Chercheur au CNRS (Institut Orléanais de Finance) et à l' OFCE. Research Affiliate au CEPR, Londres. Adresse: OFCE, 69 quai d'Orsay 75007 Paris.

(1) "The preparation of the changeover to the single european currency" (interim report), January 20 1995.

(2) Or (if they decide to sterilize the impact of their interventions) renouncing to the amount of seigniorage that they would collect if they were not part of the arrangement.

(3) See Kenen's analysis of the Emminger letter (Kenen: 1995).

(4) Goodhart questions the efficiency of Article 104 arguing that indirect monetization could happen if central banks colluded with governments to repurchase public debt from commercial banks. However, provided that national central banks are independent from national governments and have a clear mandate for price stability, this problem should not arise.