FLEXIBILITY AND ENHANCED COOPERATION IN EUROPEAN SECURITY MATTERS: ASSETS OR LIABILITIES

FLEXIBILITY AND ENHANCED COOPERATION IN EUROPEAN SECURITY MATTERS: ASSETS OR LIABILITIES?

© Institute for Security Studies of WEU 2000. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photo-copying, recording or otherwise without the prior permission of the Institute for Security Studies of WEU. ISSN 1017-7566

The present Occasional Paper collects some of the materials produced for the Seminar on Flexibility and Enhanced Cooperation in European Security Matters: Assets or Liabilities?, that the WEU Institute for Security Studies organised on 22-23 October 1998 in Paris. More specifically, the reader will find the background paper that was distributed in advance to the participants, the brief general introduction to the seminar proper, the main papers delivered on that occasion, a couple of interventions in the debate, and the conclusions. The annex contains the relevant articles of the Amsterdam Treaty - on CFSP and 'flexibility' - in their consolidated version.

By publishing and distributing all of this as an Occasional Paper without much editing, we intend to disseminate some food for thought in the wake of the recently revived discussion over European defence, and in anticipation of the entry into force of the new Treaty. We are grateful to the contributors for having entrusted to us their contributions and accepted that, for expediency's sake, they be distributed in a form that is not customary for the Institute.

Background

Since it was inserted in the Maastricht Treaty on European Union (TEU), the common foreign and security policy (CFSP) - the EU's second 'pillar' - has become subject to intensive academic, political and public attention, and criticism. A tangible "expectations vs. capability gap" has affected its implementation as well as its reputation, while other "policy providers" - UN, NATO, OSCE, but also various 'coalitions of the willing', each one with EU member states at the forefront - have more or less adequately met the challenges with which they were confronted ⁽¹⁾ . To a certain extent, the same may be said about the Western European Union (WEU), which has just celebrated its 50th anniversary ⁽²⁾ .

After the signing of the Amsterdam Treaty, the only policy areas in which a clear European identity fails to emerge and in which going beyond strict intergovernmentalism seems a hardly attainable goal are precisely those that CFSP and WEU address. This does not mean, of course, that the EU as such has no international presence at all, nor that it cannot claim to be an international actor of some weight in its own right: as a matter of fact, its common trade and aid policies - as key components of the first 'pillar', they are carried out by the Commission and shape the EC's proper "external relations"- represent important political levers on the global scene and have already lead to the conceptualisation of the EU as a "civilian power" ⁽³⁾ . In addition, enlargement of the EC/EU has always been B and is even more so now B also a powerful foreign and security policy tool in that it is designed to project prosperity, stability and a single set of values and rules beyond the original 'family' core ⁽⁴⁾ .

Yet it is apparent that they fall short of a proper common foreign, let alone security policy: institutionally, trade and aid, humanitarian missions, external relations and CFSP proper are being dealt with by separate bureaucratic entities (as well as by different EU Commissioners); academically, there is no consensus over considering all them as components of one and the same policy area. Besides, even along the long and winding road leading from EPC ⁽⁵⁾ to CFSP many opportunities to strengthen and upgrade the existing intergovernmental cooperation (and to make it more consistent with the existing common trade and aid policies) were missed, neglected or inadequately addressed by the EC and the EU member states - be it in the Middle East, vis-a-vis the apartheid regime in South Africa or, more recently, in former Yugoslavia - and have consequently led to widespread disappointment and frustration.

As a partial response to such difficulties, lately - that is, before and during the negotiations that led to the Amsterdam Treaty - some analysts and policy-makers floated the idea of resorting to some form of flexibility or enhanced cooperation within the European institutional framework in order to allow more effectiveness and visibility in the second pillar and to make "Europe" a more respected international actor.

This paper will provide a brief survey of the debate, a summary of the IGC negotiations, an assessment of their outcome and implications for CFSP and WEU, and a brief analysis of the present state of affairs - all of them sub specie of flexibility and/or enhanced cooperation.

The debate about how institutionally to reconcile and manage heterogeneity within the European Community/Union (EC/EU) - with a view to the dual challenge of "deepening" and "widening" - is hardly new. From the publication of the Tindemans Report in 1975 until the early 1990s, though, the supply of quality literature and convincing arguments on the subject was scarce and occasional. Only after the Maastricht Treaty was ratified and the CDU/CSU parliamentary group released its controversial paper 'Reflections on European Policy', in September 1994, did the discussion take off again, giving way to an avalanche of new visions and concepts across the continent: Europe, it was suggested, should go multi-speed and for an integration échelonnée ; it should become two-tier, multi-track, variable-geometry, or à la carte; it should be built around a hard core, or take the form of concentric circles. Scholars and political leaders competed in the coinage of terms that clearly entailed different visions and goals, which in the end made the whole discussion fairly confusing ⁽⁶⁾ .

At least until the conclusion of the work of the so-called Reflection Group (July-December 1995), however, three main definitions of a more flexible, differentiated EU may be singled out with some clarity, at least as an effort to better understand some of the implications of the whole discussion: the multi-speed, the variable geometry, and the à la carte Europes. Roughly speaking, again, the three could be pegged, respectively, to time, space, and subject matter. A multi-speed EU, for instance, is aimed at a mode of integration whereby the pursuit of common objectives is driven by a group of member states which are both willing and able to 'deepen' their integration in some policy areas (the underlying assumption being that the others will follow later). Such vision is positive in that, although admitting differences, the member States maintained the same objectives, which would be eventually reached by all members in due time, and were all guaranteed full participation in the related decision-making processes. The second main concept - variable geometry - admits instead substantial differences within the main integrative structure by allowing lasting or permanent separation between a core of countries and lesser developed integrative units. Such Europe differentiated by space, therefore, goes further in institutionalising diversity than a mode of integration differentiated by time: whereas the latter would define and maintain a full range of common objectives and goals, differentiation by space takes a less ambitious approach and acknowledges that, due to its internal diversity, Europe will and should organise itself, at least for the foreseeable future, around various integrative units. Finally, by definition, the culinary metaphor of a Europe à la carte would allow each member state to pick and choose, as from a menu, the policy area in which it would like to participate: at the same time, a minimum number of common objectives would be maintained. In principle, therefore, all countries would be in the first circle, in which they could each choose their subject matter/area of participation - be it social policy, monetary policy, or defence policy - and opt out from the others ⁽⁷⁾ .

Of course, things are not so simple and clear-cut: even the CDU/CSU proposal, for instance, called for a 'multi-speed' or 'variable geometry' Europe, without realising that the two concepts are rather different ⁽⁸⁾ . It is arguable that the multi-speed and the à la carte approaches are at the two opposite extremes of the spectrum of differentiated integration, whereas - in its ambiguity - variable geometry exemplifies the middle ground (or grey area) in between. The main difference between variable geometry and multi-speed is the degree of common objectives involved. Variable geometry has to take place outside the acquis communautaire and - as opposed to à la carte - somewhat implies various forms of deeper integration outside the normal decision-making framework of the EU. Multi-speed, on the other hand, aims at the most ambitious acquis and avoids any form of differentiated integration outside the community structure.

In terms of policy, and more concretely, transitional periods and temporary derogations- often related to accession agreements, harmonisation of VAT, or similar issues - could already be considered the most obvious examples of a multi-speed integration already in place. Moreover, the Maastricht Treaty introduced a very important multi-speed element into the implementation of EMU: as a matter of fact, capability (as measured against common and unanimously agreed convergence criteria), willingness and a precise timetable jointly make for a quintessential case of multi-speed Europe, in which it is up to each member State to choose the most appropriate means to achieve such common goals. Indeed, EMU itself has become more or less explicitly the cornerstone - or rather the underlying term of reference - of any controversy over enhanced integration ever since ⁽⁹⁾ .

Within European institutions there have been also examples of variable geometry. In the area of security policy, of course, this was illustrated by the WEU itself, and by the Eurocorps, Euromarfor, Eurofor and other bi/multinational forces. And in the sphere of the initially fledgling 'third pillar' (justice and home affairs), the Schengen Agreement could be considered a good example of a conglomeration of states pursuing deeper integration within a separate unit through forms of opting-in (rather than opting-out). Finally, the Maastricht Treaty itself enshrined a number of opt-out clauses that could be situated somewhere between à la carte integration and variable geometry: on EMU (United Kingdom, Denmark), on the common defence policy (Denmark), and on the Social Charter (United Kingdom). Some residual evidence of opt-outs can also be found in the accession agreement signed in 1994 by Sweden (the 'snuff' tobacco exemption).

This is only to say that, fuzzy as it may have appeared, the 1994-95 debate had a bearing on many forms of European integration that were not purely theoretical, nor entirely untested. Indeed, Wolfgang Schäuble, Edouard Balladur and John Major - as well as their academic counterparts - probably had something fairly specific in mind when they outlined their respective blueprints for a more differentiated Europe ⁽¹⁰⁾ .

The Reflection Group preparing the IGC published its final report in December 1995, at the end of the Spanish EU presidency. Although the Group's progress report, too, displayed a certain inconsistency in the use of terminology related to differentiated integration within the EU, its report outlined a clear vision of its limits and future possibilities. In essence the Group, chaired by Carlos Westendorp, rejected any formula which could lead to an à la carte Europe. It maintained that more flexible solutions - mainly multi-speed - could be used if the following criteria were met: a) differentiation should only be allowed as a last resort and temporarily; b) those who are willing and able should not be excluded from participation in a given action or future policy; and c) when allowing differentiation, the acquis communautaire and the existing single institutional framework should be preserved and respected. The report also pointed out that the degree of differentiation admissible varied according to the pillar in question, and also between the present member states and those acceding in the next enlargement(s). In other words, while derogations would not be allowed in the first pillar if they could jeopardise the internal market, the CFSP and some third-pillar issues may enable a greater degree of differentiation. All this meant that multi-speed integration continued to be the name of the game - inasmuch as it ever has to be played - while à la carte integration was rejected outright. Variable geometry formulas, on the other hand, were not addressed explicitly by the Group - which may help explain the invitation to do so made by Chancellor Kohl and President Chirac in an open letter dated 5 December 1995.

The Franco-German move also influenced the debate by introducing a new term - coopération renforcée/ enhanced cooperation - into the political vocabulary of the IGC, and by suggesting that its qualified insertion into the revised TEU should be properly considered. From then on, therefore, the political and academic debate was centred mainly on enhanced cooperation and coupled - again, a little confusingly - with 'flexibility'. Indeed, each term concealed different, even divergent, views on the future of European integration. One of these - a more centralising view - considered 'enhanced cooperation' as a half-way house towards bringing more competences and activities within the EU, and as a way of making greater use of Community procedures and institutions. The other - a more decentralist view - assumed that Europe could best develop around a limited core group of activities, while in other areas allowing different groupings of member states 'flexibly' to pursue different approaches and to use different procedures and institutions suited to the policy in question. As a consequence of such terminological ambiguity - partly inevitable and perhaps necessary, at that stage of the IGC negotiations, partly unhelpful to the ensuing discussion - the German Chancellor, the French President and the British Prime Minister all endorsed flexibility, although each meant something different by it.

However ambivalent its acquis linguistique, the IGC was eventually set in motion in Turin, in late March 1996, and went on through the Irish and Dutch presidencies before coming to an end at the European Council in Amsterdam, in June 1997, immediately after the British (and the French) parliamentary elections. As mentioned above, the political impulse was established in two joint Franco-German letters published before the IGC (i.e. the letters by Kohl/Chirac of 7 December 1995, and by Kinkel/de Charette of 27 February 1996). The European Council of Turin provided the mandate for examining enhanced cooperation/flexibility in the IGC, and it was clear from the beginning that it would be one of the most difficult and sensitive areas of discussion. Throughout the Conference, a total of twenty-two documents were submitted on the matter: apart from those released by the successive presidencies, France and Germany (jointly), Italy, Portugal and Greece (separately) all submitted their own texts and proposals. In addition, a number of non-papers, such as the 'Ten Commandments of Flexible Integration' by the Finnish delegation, were circulated among the participants.

Without dwelling on too many details and technicalities, suffice it to say here that the evolution of the discussion was typical of any new concept developed in any intergovernmental conference: first, the idea was launched; second, the concept was defined; third, a draft article was provided; and finally, the latter was subjected to interpretation and negotiations. In particular, the Italian presidency's report on the state of play raised the idea of a general flexibility clause - for purely practical reasons, from now on the term 'flexibility' will be used as synonymous with enhanced cooperation - supported by three specific flexibility clauses for each pillar (June 1996). Incidentally, the move on the general clause was probably dictated, at least partly, also by the obstructive behaviour adopted at the EU level, in the Spring of 1996, by the United Kingdom in retaliation for the embargo on British beef issued by the Commission to prevent the "mad cow" disease from further spreading. Flexibility, in other words, was also seen as the only way out of the institutional paralysis a single government proved able to generate.

At any rate, in the ensuing negotiations different options were floated on the mechanisms designed to 'trigger' flexibility in each pillar: for instance, qualified majority voting (QMV) in the first, unanimity in the second, both - alternatively - in the third (the Franco-German memorandum of 17 October 1996 stressed that no member state should have a veto right). In January 1997 Italy submitted a draft article for the second pillar, whereby all forms of flexibility related to defence required the consent of all WEU members ⁽¹¹⁾ . Later on, though, the Dutch presidency voiced doubts as to the necessity of an enabling clause in the second pillar and, in any case, the final draft prepared for the Amsterdam summit envisaged unanimity as its trigger mechanism, whereas QMV was deemed sufficient for the first and third pillars. This turned out to be an important warning signal: in the final, hectic stage of the Amsterdam negotiations the entire flexibility clause in the second pillar literally disappeared from the table - most participants claim, however, that general agreement on its inappropriateness for CFSP had been reached much earlier - and was dropped in favour of "constructive abstention". By and large, it may be argued that the arrival on the European stage of a new British government officially less hostile to constructive cooperation with its EU partners - along with the consolidation of a less obstructive political leadership in Greece⁽¹²⁾ - contributed to rest the case and to lower the call for a 'strong' general flexibility clause. All this made also possible to address more traditional concerns about the preservation of the principles of the Treaties and the single institutional framework of the Union and the use of flexibility only as a last resort (Art.K.15/ now Art.43 consol. TEC).

In the end, the Amsterdam Treaty endorsed three basic forms of flexibility, although - significantly enough - the term as such completely disappeared from the final text:

1) enabling clauses , i.e. the mode of integration which enables willing and able member states to pursue further integration, subject to certain conditions set out in the treaties, in a number of policy areas within the institutional framework of the EU. Examples include a general flexibility clause to be inserted as a new Title to the common provisions of the TEU, as well as clauses specific to the first pillar (Art.5a/ now Art.11 consol. TEC) and the third pillar (Art.12/ now Art.40 consol. TEU).

2) case-by-case flexibility, i.e. the mode of integration which allows a member state the possibility of abstaining from voting on a decision by formally declaring that it will not contribute to the decision, whilst at the same time accepting that the decision commits the entire EU. This so-called "constructive abstention", therefore, is both a decision-making mechanism and a trigger mechanism for flexibility. As already mentioned, it applies only to the second pillar (Art.J.13/now Art.23 consol. TEU) and is designed to offset the eleventh-hour 'disappearance' of a specific flexibility clause;

3) pre-defined flexibility, i.e. the mode of differentiated integration which covers a specific field, is pre-defined in all its elements (including objective and scope) and is automatically applicable as soon as the Treaty enters into force. It is primarily established in protocols and declarations related to the previous third pillar, and affects specifically and explicitly Denmark - which is inside the Schengen "space" but with an opt-out for the rest of the third pillar - the United Kingdom and the Republic of Ireland, both outside Schengen but with an opt-in for the rest of the third pillar.

All in all, and with specific reference to the terminology used in the pre-IGC debate, the first form is the closest thing to a new multi-speed framework, but with plenty of constraining conditions, which have led some advocates of enhanced cooperation to complain about a "strait-jacket" being imposed on any future grouping of willing and able ⁽¹³⁾ . At the opposite end, the third form is close to a Europe à la carte in which the orders have already been taken: all the customers have opted for the basic menu on offer at a reasonable price, with some of them (vegetarians maybe, or religious-minded) getting ad hoc variants without any extra-charge; if proven satisfactory, the whole arrangement may quickly become the staple diet. Finally, the second form dramatically shows the limits of institutional provisions in tackling issues and policy areas where fundamental disagreement persists among member states - e.g. about what political responsibilities should be attributed to the EU, about the future of European defence, and about how it should be configured with respect to NATO - and where incentives for enhanced cooperation are either comparatively low (other organisations already in place do the same job more effectively) or unevenly distributed between security 'providers' and 'consumers' ⁽¹⁴⁾ . Given also the de facto postponement, at Amsterdam, of the decision on the integration of WEU into the EU - a detailed three-stage proposal to this end had been jointly submitted to the IGC by Belgium, France, Germany, Italy, Luxembourg and Spain in March 1997 ⁽¹⁵⁾, and a similar prospect had been suggested in the IGC recommendations of the European Parliament - it can be argued that the present state of affairs paves the way, more or less directly, to variable geometry as the only possible form of enhanced cooperation in the fields of security and defence.

Art. J.4 of the Maastricht Treaty, concerning CFSP, refers to "the eventual framing of a common defence policy, which might in time lead to a common defence", and looks at WEU as the institution which, as "an integral part of the development" of the EU, may be "requested" by the EU to "implement decisions and actions of the Union which have defence implications". Very little use has been made so far of such a provision ⁽¹⁶⁾ . Moreover, the fact that the three new EU member states who joined after the ratification of the TEU (Austria, Finland and Sweden) were all neutrals/non-aligned, albeit with different traditions and identities, and that all eventually opted (as non-NATO) for "observer" status within WEU's institutional architecture has further differentiated the EU/WEU membership chart. In actuality, Austria, Finland and Sweden all have remarkable records in terms of participation in multilateral peacekeeping and similar operations (now categorised in WEU as "Petersberg-type" missions) under the UN flag ⁽¹⁷⁾ . However, they have significantly added to the already existing mismatch between full memberships of the EU, WEU and NATO. And the actual dynamics of the IGC negotiations showed how the initial purpose of allowing some kind of enhanced cooperation among member states in policy areas where unanimity was compulsory - the citing of particular instances in which some national governments had blocked agreement on CFSP issues weighed heavily in the debate - eventually yielded to more traditional solutions.

As already mentioned, the only loophole left in the Amsterdam Treaty for undertakings that are not strictly unanimous - once that, much to Italy's chagrin, not only the flexibility clause but also the extended use of QMV in the second 'pillar' were thrown into the IGC dustbin - is the "constructive abstention" clause (Art.23 consol. TEU), in itself a rather ambivalent device. In fact, much as it makes good sense that a reluctant member State - when no "important" and "stated reasons of national policy" call for a formal veto (this being the 'emergency brake' inserted in the Treaty at the eleventh hour) - simply refrains from action without blocking a sizeable majority of the others, how far can such a 'consensus minus X' formula be stretched without undermining the credibility of the decision and its implementation? As a matter of fact, here may arise problems of both quantity and quality of the abstentions, let alone the fact that "constructive abstention" is not applicable to decisions with military and defence implications anyhow (Art.23 consol. TEU). On the one hand, in fact, if the abstainees make for more than one third of the 'weighted' votes in the Council, the decision is not adopted. On the other, it is not irrelevant who abstains on what decision.

Of course, it is somewhat arbitrary to draw up virtual scenarios in the absence of specific cases and configurations: it is likely, however, that a common action with 'soft' military implications from which, say, France and the UK decide to abstain would appear less credible than one from which, say, Ireland and Finland decide to abstain. Plausibly, in fact, the former would not take place at all - or would take place outside the W/EU framework - and the latter may become part of a broader politico-military framework involving e.g. the UN and/or NATO. Moreover, abstentions should be measured also against the 'theatre' they refer to: if Scandinavian countries abstain from a decision affecting the Baltic Sea area, for instance, the credibility of that same decision - let alone its implementation - will presumably be lesser than if the abstention comes from, say, Portugal or Greece: in other words, there should preferably be some consistency between voting behaviour and actual engagement in the target area. Besides, it will turn increasingly difficult to achieve unanimous consent and commitment from 15-plus member states to carry out such diverse "Petersberg-type" missions as those now explicitly envisaged also by the Amsterdam Treaty (Art. J.7/now art.17 consol. TEU). Finally, the abstention clause might have a much less constructive impact on financial arrangements, if it leads a country to abstain for purely financial reasons. The costs of missions approved through "constructive abstention", in fact, will be borne by the participating member states in accordance with the GDP scale - and not by the Community budget, unless the Council decides otherwise - with the exemption of those who opt to abstain. All this, to put it bluntly, may pave the way for new forms of free-riding in foreign policy and may occasionally turn "destructive", rather than "constructive", for CFSP ⁽¹⁸⁾ .

The transfer of a similar device to the WEU Treaty or to WEU practice - as recently suggested by some WEU member states, most vocally by France - could generate similar problems while, of course, contributing to the solution of others ⁽¹⁹⁾ . On the one hand, the WEU has to a certain extent already become a multi-tier (albeit not a multi-speed) organisation: since 1995, apart from a core group of 10 full members, it has acquired successive 'circles' of companions - 3 associate members (NATO-only members Norway, Iceland and Turkey), 5 observers (Denmark plus EU-only members Ireland, Austria, Finland, Sweden), 10 associate partners (all the Central European EU candidates) - but it increasingly works either on a 18 or on a 28 basis: at 18, obviously, when matters related to EU and NATO are addressed. On the other hand, the kind of missions WEU has carried out so far have been particularly low-key, low-risk and low-cost. Thus there has been no need so far to resort to any form of abstention (irrespective of the financial arrangements) nor to open a debate on any form of 'flexibility'. At the same time, however, and in perspective, the present Verfassung of WEU already envisages a dual system of guarantees and potential commitments: the inner circle of 10 full members of both NATO and EU clearly entails 'hard' mutual security guarantees (Art.V of the modified Brussels Treaty), and has stringent defence implications; the outer circle gradually encompasses the whole company of 28 and, irrespective of present and/or future memberships of the EU and NATO, is already potentially available to carry out non-Art.V, "Petersberg-type" missions. Logically, it is only the latter circle's size that could call for some form of "constructive" opting-out, should the WEU framework be used for multilateral operations.

Article 17 consol. TEU now defines WEU as "an integral part of the development of the Union" in that it supports the EU "in framing the defence aspects of the common foreign and security policy". Moreover, it reads, the EU "will avail itself of WEU to elaborate and implement decisions and actions of the Union which have defence implications", and WEU will be directly involved in the setting up and the subsequent activity of the forthcoming Policy Planning and Early Warning Unit (PPEWU) envisaged by a Declaration to the Amsterdam Treaty, which in turn will be under the responsibility of the equally newly appointed "High Representative" (most commonly referred to as "Mr./Ms. PESC"), expected to embody CFSP and to act as the mover and shaker inside the Council. On paper, therefore, now that "Petersberg-type" tasks are explicitly included in the policy sphere of CFSP, the enhanced 'agency'-role that Art.17 attributes to WEU in the framework of the "progressive" establishment of a common defence policy provides the functional connection for closer cooperation and interaction between the two institutions.

From closer up, however, the picture looks more intricate, especially as far as the decision-making procedures are concerned. In principle - as regards the management of Petersberg-type peace support operations - European countries currently have a broad range of institutional options at their disposal:

In actuality, though, all these options have to be measured against the same countries' political will and the general attitude of the United States, especially at the higher range of "Petersberg-type" operations. There is far less debate in Europe than frequently imagined about the proposition that in any crisis where the U.S. declares itself willing to play an active role, NATO will continue to be the main forum for politico-military deliberations. Even in the 1998 Kosovo crisis, the United States very early on demonstrated its interest in playing a key role in its management, thus automatically transferring the primary locus of planning and deliberations to NATO. At the same time, however, the American commitment to participate in European crisis management is neither automatic nor unconditional. As shown in the early stages of the Yugoslav crisis in the early 1990s and in the 1997 Albania case, we have already witnessed occasions on which a number of European countries saw their national security interests directly jeopardised and on which the United States declared itself unwilling to intervene. Furthermore, in recent times the gap between current Clinton administration policy on American involvement in international security matters and the domestic political debate about these issues in the U.S. public at large, and in the U.S. Congress in specific, has grown as wide as it has ever been.

On top of that, there is a growing gap also in the way in which Americans and Europeans conceptualise and implement their respective approaches to conflict management: suffice here to mention the so-called revolution in military affairs (RMA, 'softwar') as compared, for instance, with the illustrative mission profiles that WEU recently submitted to NATO for planning purposes - namely conflict prevention, assistance to civilians, guarantee and denial of movement, imposition of sanctions, containment and separation of parties by force. Such gap does not rule out either a sound division of labour or a partial overlap between different institutions: the present debate on NATO's new strategic concept is a good case in point. The whole setting up of CJTFs, however, is made even more complicated - politically as well as operationally - by the fact that NATO capabilities cannot easily be separated between collective defence and crisis management functions. Yet it must be said also that recent developments in the restructuring of national armed forces - most notably in France, Spain, Italy, even Sweden (Britain had started much earlier) - hint at a reorientation towards inter-army projection of force and interoperability (rather than territorial defence or deterrence) that, along with other developments in the field of defence industry and joint arms procurement, looks promising for CJTFs and WEU-led missions to come ⁽²⁰⁾ . Finally, it goes without saying that Europe does not have the United States' ambitions for global military power projection, so that any crude comparison in this field - from spending to technology - would be somewhat unfair and not to the point.

What are the consequences of these new trends for European crisis management ? From a European point of view, the most likely and preferred military option in a crisis involving European security interests in or around the continent today remains the NATO option with American participation: either a full-fledged NATO operation, or a NATO-led CJTF-operation (as allowed for by the 1996 Berlin Declaration) with the participation of third countries. As already said, all European countries de facto agree that an operation with the direct involvement of the United States is to be preferred over an operation without it. In addition, such primacy was explicitly acknowledged in both WEU declarations that were attached to the 1991 Maastricht and the 1997 Amsterdam Treaties, which called NATO "the essential forum for consultation" and "the framework in which [the members] agree on policies bearing on their security and defence commitments under the Washington Treaty". This does not mean, however, that NATO will always play the leading role in every conceivable operation. In potential crises where Europe and the U.S. have asymmetrical (if not contradictory) interests, it looks prudent and farsighted to develop some residual Europe-only options in case the United States is unwilling (for whatever reason) to participate militarily. Here is where W/EU comes into focus, particularly if "Europe" does not intend to remain a purely "civilian" power B even in its 'stronger' version, which encompasses all sorts of diplomatic means as well as trade and economic sanctions ⁽²¹⁾ - and where, given the above mentioned difficulty to reach unanimous consensus and commitment among 15-plus members, the need for more 'flexible' arrangements is most acutely felt.

So far political scientists, and especially international relations theorists, have devoted scant and only intermittent attention to the role of 'cores' and 'clubs' in international organisations and, more generally, in defence and security matters ⁽²²⁾ . The existing academic literature is therefore of limited help in envisaging viable ways to overcome or sidestep the institutional blockages that affect European crisis management. Moreover, even under the terms of the new EU Treaty, the second pillar cannot by itself produce a 'core' group committed to a common security and defence and cannot therefore bring about an effective multi-speed CFSP: the impulses the EU will be able to give to WEU will presumably be varied but limited in scope and intensity, and may in any case still require a concomitant unanimous decision by WEU itself (and by NATO, if the CJTF concept or some 'borrowing' of assets are drawn into play). By contrast, WEU is legally entitled to decide autonomously on a security and defence action (including non-Art.V missions) without a concomitant decision of the EU, although many efforts are being made in order to fine-tune the (in)decision-making WEU/EU "flow chart" and to improve the coordination, in particular, of the rotational presidencies of both organisations ⁽²³⁾ . Which means, too, that it is ever more unlikely now that WEU may go it alone, since both its inner and outer 'circles' will in practice refer to EU and/or NATO (as well as UN and OSCE). At the same time, WEU currently is the only official link between EU and NATO, which still gives it a specific function among European and Western institutions: in the critical early stages of a crisis, during which the Europeans would presumably be struggling to achieve a consensus on possible courses of action, the WEU Council still provides the only available institutional framework for military-political discussion, while the actual military option-generation and preliminary planning can be set in motion by the WEU's operational components ⁽²⁴⁾ . At times, however, precisely such 'interface' role of WEU has allowed for some cases of "forum-hopping", that is, a policy (or practice) of deferring or even boycotting decisions and actions by more or less systematically 'shuttling' them between different bodies and fora.

Finally, the uncertainties that still linger on the actual use of the CJTF concept make it rather unlikely that Europe as such - within the framework of its present institutions - will be able to decide, plan and carry out autonomously an action in this field unless a new political landscape, the accession of new members and other developments - first of all, an external shock and/or a new crisis in the vicinity - substantially change the key players' attitudes and incentives.

For some evidence, one only has to look at what happened with the 1997 Albanian crisis, although it must be taken into account that it climaxed during the final stage of the IGC negotiations (and shortly before the British elections of May 1997). On the one hand, the lesson of Bosnia had been learnt, and the international community encouraged some form of external intervention at a relatively early stage. On the other hand, once the United States made it clear that, contrary to Bosnia, it would not play any direct role in Albania, the field was open for "Europe" to act, and the formula that was eventually adopted belied most of the commitments made earlier in this domain: the EU failed to reach unanimous agreement on an intervention in Albania under the existing CFSP provisions - the Germans and the British, in particular, expressed their reluctance to take action - while the WEU Council did not consider the possibility of taking independent action until it was confronted with the request to play a minor role on the ground, i.e. to send a reduced multinational advisory police element (MAPE), in the light of the experience previously acquired in Mostar.

In the end, a very mixed "coalition of the willing" - Italy, France, Spain, Greece, Romania, Slovenia, Turkey, plus minor contributions by Denmark, Austria and, towards the end, Belgium - set up "Operation Alba" with a UN/OSCE humanitarian mandate and under Italian "framework/lead nation"-type leadership ⁽²⁵⁾ . The mission was therefore undertaken outside any specifically European institutional framework: it lasted a relatively short time (as compared with IFOR/SFOR) - April to August 1997, including a short extension of the original mandate - and proved relatively successful, given the initial difficulties on the ground. The fairly positive outcome may also explain why the W/EU governments who had resisted a common action (by simply refraining from it, under Maastricht Treaty rules) seem to have had second thoughts and now, with the benefit of hindsight, regret not having made 'Alba' the first test-case - ahead of the entry into force of the Amsterdam provisions - for an effective CFSP-WEU joint action.

Of course, it is difficult to assess now whether an 'Alba'-type operation could have been set up and carried out under the novel CFSP provisions of the Amsterdam Treaty. As a matter of fact, it was also the controversial nature of the final negotiations over the Treaty that led e.g. Britain to undermine the hypothesis of a EU common action. Yet if there is one lesson, among others ⁽²⁶⁾ , to be drawn from Alba, it probably has to do with flexibility, enhanced cooperation and variable geometry.

On the one hand, it is arguable that the present institutional provisions - given the substantial lack of convergence on crucial political issues, along with different geopolitical priorities and with only occasional bouts of willingness to act on the part of W/EU member states - are not sufficient to provide solid ground for the future evolution of CFSP. With some qualifications, the "constructive abstention" clause of the Amsterdam Treaty may help overcome some obstacles, but - as argued above - should be handled with great care and, by definition, on a case-by-case basis. Furthermore, it is not applicable to defence matters anyhow.

In a sense, once again, only a successful start of the final stage of EMU might provide new momentum for strengthening cooperation and solidarity among its core members on an increasingly wider range of policies, therefore (hopefully) linking more tightly the first and the second pillar and eventually producing positive results for CFSP as well. Yet, in spite of all spillover theories and expectations, there still is no guarantee nor any evidence that this will happen in the foreseeable future. For one thing, it is true that EMU's likely initial configuration at 11 - instead of '5 plus', as envisaged in the early stages of the debate over a 'core' EU - has eventually had a reassuring effect in some countries with respect to other future 'flexible' arrangements: in the first as well as in the other 'pillars', in fact, it will become increasingly important to build large coalitions of interested actors rather than to rely on bi- or trilateral axes as the main driving force. It is also true, however, that the only tangible spin-off of EMU so far has been its capacity to exclusively concentrate and absorb the minds and actions of policy-makers - to the detriment of other policies, most notably CFSP. Finally, it goes without saying that a major crisis of EMU would fatally affect every other common undertaking and put at risk the prospect of Political Union itself. Conversely, of course, a successful management (and enlargement) of the Euro-core could boost further integration in other policy fields by consolidating the perception of common external interests.

On the other hand, 'Alba' has proved that coalitions of the willing built around a common objective and whose scope is limited can work quite effectively, at least as far as they apply to low-scale peace support operations. As such, it has set a useful precedent on which to build in the future, although it has hardly solved the issue of the international legitimisation of such 'ad hoc' groupings, nor has it significantly strengthened the case for institutionalised multilateralism. At any rate, Alba certainly was a Petersberg-type mission, with the involvement of a limited number of countries: in a way, it also was a quintessential case of enhanced cooperation with a different label, albeit 'enlarged' to non-EU and non-NATO members. Even such openness to outsiders and 'pre-ins', however, could be considered a promising experience in the light of future developments in this field, provided it will not produce a permanent multi-tier framework with unequal obligations and unequal rights for participants (a problem, incidentally, that may soon involve not only NATO and/or EU candidates but also such 'post-neutral' States as Austria, Finland, and Sweden). In fact, it should not be forgotten that enlargement is in itself a policy that significantly affects the overall prospects for peace and stability in Europe: such enlargements de facto or ahead of time as Alba or, to a certain extent, WEU's Kirchberg Declaration (1994), if properly managed, may well reinforce already existing trends and anticipate further cooperation and integration. And, finally, one could go as far as to say that after all - given the use of NATO's Partnership for Peace operational standards and codes on the ground - Alba was a sort of CJTF-type mission ante litteram.

If variable geometry has any meaning, then, it is related precisely to such a configuration and refers to policies that are:

European governments may still lack the method, but the general idea is there. Even the drafters of the Amsterdam Treaty seem to have guessed as much when they wrote that Article J.7. (Art.17 consol. TEU), with all its constraints and tortuous formulations, "shall not prevent the development of closer cooperation [italics added] between two or more member states on a bilateral level, in the framework of the WEU and the Atlantic Alliance, provided such cooperation does not run counter to or impede" CFSP as previously outlined. Of course, if the cooperation among Europeans aims to go beyond intergovernmental mini-lateralism and purely reactive contingency planning, it will have to move from 'ad-hocery' - with all its pros and cons - to a more institutionalised multilateral framework. To take root and to win the day, however, such framework should be able to strike the right balance between institutional and operational flexibility, political credibility, and overall effectiveness.

In sum, enacting and testing enhanced cooperation outside too rigid institutional settings, yet with:

- with a view to stabilising peace support in the medium-long term and to reinforcing civilian reconstruction in conflict-plagued areas (the latter being, with the benefit of hindsight, Alba's main shortcomings),

could well turn out to be the essential prerequisite to giving more legitimacy to the second 'pillar', to accumulating a sort of acquis securitaire, and to preparing the ground for the insertion of such forms of flexibility into the EU (and WEU) framework at a later stage - which, incidentally, is what has just happened with the third pillar, that has eventually been much more 'communitarised' than it seemed originally possible, given the quintessentially intergovernmental and 'flexible' shape that it had taken from the outset. ⁽²⁷⁾

At the same time, precisely the experience of the 'Schengen' group shows that these forms of variable geometry eventually need to be inserted, more or less explicitly, into a political structure capable of providing leadership and responsiveness beyond the lifetime (and the occasional convergence of interests) of "coalitions of the willing" or regional 'directoires' . Such political structure exists for the U.S. foreign policy and, to a lesser extent, for NATO. Yet it does not exist, or rather it is not sufficiently nor adequately developed, in the W/EU compound, and the Amsterdam Treaty per se has not particularly fostered its growth. It remains to be seen whether the hypothesis B recently floated, for instance, by a British Foreign Office non-paper - of a EU "fourth pillar" centred on defence and encompassing WEU can provide a viable one, and whether such a structure would be able to strike a satisfactory balance between all the potentially contradictory factors described above. At any rate, since Germany's take-over of the joint EU/WEU rotational presidency in January 1999 (due to coincide with the entry into force of the Amsterdam Treaty provisions), the European Council will have a one-year spell to reassess the state of play in order also to possibly decide - by unanimous vote, of course - over their integration/merger without calling an Intergovernmental Conference.

Introduction

"Europe" already has a variable geometry. There is the Europe of the single market, encompassing the 15 member States, plus the EEA (Norway, Iceland, Liechtenstein) and, increasingly, the candidates for accession. Moreover, there is the Europe of EMU, encompassing the 11 States of the first wave, the 2 of EMS-II (Denmark and Greece) and, on the outer circle, the opter-out Britain and the unqualified stayer-out Sweden. Then there is the Europe of Schengen, encompassing the 15 minus the two opter-outs (Britain and Ireland, with Denmark as a possible opter-in), plus Norway and Iceland (through the Nordic passport union). Each 'space' is linked to a specific 'regime' and has its own acquis. Not so, however, for security and defense: there certainly are many overlapping 'spaces' (WEU, NATO, the Partnership for Peace program, the EAPC Council), yet 'Europe' as such has no distinctive regime nor any specific, tangible, single acquis to be incorporated or respected by existing or future adherents: if there is any, it is either very weak and fragmented (WEAG, OCCAR), or mainly linked to NATO.

When a major crisis arises, then, its management necessarily involves a whole set of institutions: EU, WEU, NATO, OSCE, the UNO. Like it or not, crisis management increasingly involves also such informal fora as e.g. the Contact Group - let alone other minor mini-lateralist groupings - as well as occasional 'ad hoc' coalitions of the willing' (e.g. 'Alba'), all usually operating outside of existing and more or less tested multilateral frameworks. In a perfect world, this would be a case of mutually reinforcing and interlocking institutions. In actual Europe, it very often shows mutually debilitating and interblocking institutions. As seen from WEU, in particular, all this often leads to sophisticated forms of "forum-hopping". In a recent joint assessment on the decision-making 'flow-chart', officials from EU and WEU acknowledged that "it appears difficult to set up a standard procedure for interlinking the decision-making processes of both organizations": in sum, forum 'shuttling', duplications of decisions, reference to other bodies (NATO, OSCE, UNO) fatally affect European crisis management. Let me give you the most striking example: until last month, art.J.4.2 of the Maastricht Treaty - the one that 'triggers' resorting to WEU - has been used only once (when the EU called on WEU for the possible organization of a humanitarian operation in the region of the African Great Lakes, in November 1996) and to no practical effect. Not that opportunities to make further "requests", on the part of the EU, or even to act independently (on the part of WEU) were not at hand over the past years: yet 'Alba' was set up and implemented by a coalition of the willing, and MAPE is a very limited police operation. Only recently - perhaps in anticipation of the entry into force of the Amsterdam Treaty, yet still one treaty behind - has the European Council intensified its recourse to WEU, albeit mainly for some pre-planning and feasibility studies: on preparing an international police operation in Albania, again, on assisting in a mine clearance operation in Croatia, and on monitoring the situation in Kosovo through the WEU Satellite Center in Madrid.

On top of all this - or rather at bottom - there is, of course, a fundamental lack of consensus among Europeans about what political responsibilities should be attributed to the EU, about the future of European defense, and about how it should be configured with respect to NATO. Here is where 'flexibility' and enhanced/closer cooperation come into the picture, although they were not part of the IGC discussion on CFSP in the first place. As we all know, in the wake of Amsterdam - its 'interface' role between the EU and NATO notwithstanding - WEU has seen its 'agency' role vis-a-vis the EU increased, at least on paper. The related provisions (constructive abstention) make for a quintessential example of 'case-by-case' flexibility which puts its main emphasis on non-participation . There is no acquis, no compulsory financial solidarity - hence the risks of free-riding and of a loss of credibility - but there also is openness, and there is a sort of potential by default : the treaty does not hamper the willing and able to act, provided they form a 'critical mass' and provided no "vital national interests" are at stake. At any rate, defense as such is not touched upon by the Amsterdam 'flexible' provisions, and WEU operates - despite recent attempts at introducing forms of consensus amenage' for non-art.V missions - on a purely intergovernmental basis.

Such institutional blockages create uncertainty, indecision and paralysis, on the one hand. On the other, they reinforce the trend towards mini-lateralism and 'ad-hocery'. In organizing this seminar, therefore, we aimed at focusing on three points:

we wanted to assess the present state of play and the way in which W/EU member States conceive of flexibility and closer cooperation in the field of security and defense;

we wanted to have a frank evaluation of the pros and cons of 'ad-hocery', both in its own merit and as a possible source of an acquis sécuritaire to be gradually inserted - Schengen docet - into the acquis communautaire proper;

finally, we wanted to ascertain whether there are ways to move beyond the present state of affairs, to exploit and/or to adapt existing provisions in order to move further ahead - incrementally perhaps, but tangibly.

At the Institute we were all thrilled when we read about a UK Foreign Office paper envisaging a sort of 'euthanasia' for WEU. The thrill became even stronger when we heard that Austria, in its capacity as acting EU rotational president, has summoned a special meeting of EU defense ministers (indeed a premiere in its own right). We are old and experienced enough to know that the devil lies in the details and that fundamental decisions may take a long time, especially in Europe and on these matters. But we think there is some reason for hope, especially if we also look at what is happening in the fields of arms procurement (OCCAR treaty), of defense industry (mergers), and in connection with the restructuring of national armed forces across Europe. All these developments - if I may paraphrase the CJTF concept - still appear separate , but not entirely separable from one another. You have been invited here to help us try and put all the pieces together.

Perspectives and likely scenarios

I was asked by the organizers to introduce in 15 minutes the perspectives and scenarios for flexibility. We all received Antonio Missiroli's very interesting working paper which touches already at many of these perspectives. Confronted with an audience of specialists, I was running the risk of repeating what you might have already written and read many times. I do have a number of ideas on the topic, but I prefer to keep them for the debate we will have today and tomorrow. (Some are detailed in an article I co-authored with Geoffrey Edwards and to be published in next issue of JCMS, especially on the essence of the system, its assessment and its significance for the European integration).

I have opted for another option: I will simply introduce a number of, at times pedestrian, reflections on the different methods of forecasting used to make previsions and predictions about flexibility, their pitfalls and respective advantages. I will also try to point at a number of questions which should be given more attention, in particular the identification of possible disruptive events.

Flexibility previsions revolving around the EU can be divided into three main categories: opinion aggregation, trend extrapolation and scenario-building.⁽²⁸⁾

The overwhelming majority of opinions among 'flexibility non specialists' belongs to the first category. These opinions seem to be built on 'common sense', on a quick and literal reading of the Treaties or, even more often, on what was generally said by the first commentators 'à chaud'.

This is the quickest - 'satisficing' rather optimizing⁽²⁹⁾ - way to have some kind of answer for, say, diplomats or journalists who are frequently reassigned and have limited time to examine flexibility track record. The fact that the issue and the underlying concepts are so complex and murky reinforce the attractiveness of this option. In a way, there is some parallelism in the fate met by flexibility and subsidiarity.

A very big majority of opinions is rather pessimistic about closer cooperation ("Flexibility is not going to work" …), more positive on predetermined and case-by-case flexibility seen as second best solution and less threatening for the Community method. Reference is made to a more pragmatic future (versus dogmatic orthodoxy)

These opinions are relatively stable because flexibility or at least closer cooperation slipped from the practitioners' current agenda.

The disadvantages induced by such a method of forecasting through opinion aggregation are well know, with for instance the introduction of biases inherent to group dynamics and to the interview techniques, to name but a few. One procedure called the 'Delphi method'⁽³⁰⁾ is designed to minimize these shortcomings but its scientific value remains highly questionable. While gut feelings or intimate conviction should not be banned altogether, opinion aggregation is the weakest form of forecasting and the recommendation is usually to opt for one of the two other main approaches, trend extrapolation and scenario-building.

Those who reason in purely extrapolative terms foresee different things about flexibility as a mode of cooperation. Indeed, if they all reckon - often implicitly - that tomorrow will be like yesterday, their respective definitions or images of yesterday diverge. In other words, they choose different points of departure or main references. Extrapolating flexibility on the basis of the EMU record, the European Armaments Agency or the European Space Agency lead to very different visions of the future. Some attempt to grasp the full picture, but this requires strong analytical skills not to end up with contradictory, confused or at best vague previsions.

People who follow that line of reasoning might not be aware of the forecast technique used, namely 'trend extrapolation'. This method, the most frequently used in forecasting, relies on an extrapolation of the development of a phenomenon over time (growth and variation), sometimes by analogy.

Let's take an example: trend extrapolation applied to flexibility and the development of a common regime for armaments.⁽³¹⁾. What about the practice here? The past trend is characterized by a multiplication of attempts with a parallel development, in chronologically order, of hybrid cooperation (i.e. between extra/intra-EU) with WEAG, intra-EU cooperation with Polarm and ad hoc extra-EU cooperation with OCAR.⁽³²⁾ The sequence here seems to be, behind a multi-track approach, the constitution of an ad hoc hard core which decides on the principles; the members of the hard core can block or delay the attempts made to develop a much wider approach involving all European states (EAA); the hard core enlarges progressively and by small batches to like minded countries, so the successive enlargements can be 'absorbed' without opening window of opportunities for the renegotiation of the regime. The likely future development could be the following: after a period of 'overlapping' and partial competition between different fora, if and when the enlarged hard core has become sufficiently consolidated, a transfer in the EU structure is likely to be discussed. This 'importation' could be made effective through the activation of the general clauses now set out in Title VII of the Treaty on European Union, in other words by the establishment of a closer cooperation for armament policy (if the system is revised and extended to the second pillar, which is not currently on the agenda but is nevertheless not impossible).

The quality of this type of previsions is variable. Their accuracy supposes for instance that the power relations (rapports de force) initially defined remain relatively unchanged. It is said to be of limited value for previsions on the development of 'emerging' processes or technologies (difficulty to anticipate the rhythm and amplitude of substitution processes, then to determine the level and moment of saturation). Can flexibility be defined as an 'emerging' - institutional - technology? The question deserves to be addressed. Supposing that flexibility is not intrinsically an invention at its first stage, it is anyway particularly difficult to identify the main trend or to evaluate the significance of an event or a concept in a process geared by so many variables (classical example of a reform dismissed as low-key and dull: the Single European Act of 1985…). Besides the multi-variate nature of the EU process, forecasting has also to take into account multiplier effects, windows of opportunities and disruptive events… or even more trickier the so-called 'self-fulfilling' or 'self-defeating prophecies'. Accurate forecasting might interfere with the unfolding of events; if it is repeated and overwhelmingly believed that 'it' is a non starter, it will be a non-starter; and vice and versa).

On one hand, it could be said that the present process is the mere continuation of the main EC/EU trend, i.e. task expansion through a strategy of enlargement. Since its origins, the EC/EU has developed through clustering of the like-minded states enlarging progressively in order to preserve the highest cohesiveness.⁽³³⁾ In that respect, closer cooperation or predefined flexibility are simply new expressions of an old practice. Now that the EU is about to absorb most of Europe, the sequential or 'outside in' strategy is not an option anymore. An alternative strategy has therefore to be found in order to secure the same benefits but through internal means, including possibly some improved closer cooperation system(s).For proponents of this line of reasoning, trend extrapolation is a relevant method to identify the contours of the future flexibility cocktail.

On the other hand, it could be argued that there were different parallel trends of flexibility across policy areas and sometimes within policy areas. The distance between these trends might be reduced by the efforts made to institutionalize and even standardize specific forms of flexibility. The evolution is likely to remain a multi-trend one. In this case, the suggestion is to use a more advanced method based on scenarios building.

If the evolution in terms of flexibility is indeed to follow a multi-track pattern whose equilibrium is dynamic and unstable, the scenarios method is more adequate. Scenarios thinking has to sketch different possibilities of evolution, alternative futures which might actualize because of so-called 'disruptive events' leading to 'bifurcation points' in the history of a process. There are at least two main advantages attached to this method. First, causal links and logical sequences have to be made explicit and can therefore be discussed and challenged more easily. Second, scenarios are not stricto sensu previsions and an author is less likely to be proven wrong. The fact that probabilities assigned to scenarios are often vague is their most common weakness. Scenarios building is also not immune from implicit preferences and wishful thinking… When linked with programming and planning, this normative dimension is may be less of a problem insofar as, in this case, scenarios are by definition, supposed to determine the best course of action.⁽³⁴⁾ At this level, the scenarios builders have also to be aware of the danger of the 'toolbox approach', that's to say restricting the scenarios to existing, tested or 'pet' instruments.

This being said, it would be interesting to revisit several contributions on flexibility scenarios, for instance, the Wallace report to the Dutch Council for Public Policy Studies, "Flying together". Various scenarios have to be envisaged: mainly centripetal; mainly centrifugal; asymmetrical with areas for which a strong centripetal dynamic will operate (EMU …) and other will remain 'immune', causing a partial deepening of variable geometry. This seminar could spend some time to discuss the underlying causal links of possible scenarios and reassign probabilities to each of them.

The seminar could also focus on possible disruptive events and organizing principles which would or should preside over the use of flexibility. Enlargement is the first disruption clearly identified as such (some practitioners speak of "institutionalized cynicism over the capacity of future MS to absorb the acquis"). The EMU failure could be another one; if the blame can be put on isolated member states or grouping(s), it would probably lead to a big boost for flexibility. There is there an invitation to be creative, an invitation to identify the proper early warning signals. As far as organizing principles are concern, the debate over efficiency and divergent rationalities is equally important and difficult. Antonio Missiroli's paper points at the difficulty to design European variable geometry according to the best combination of tasks, actors, instruments (cf. study of the comparative advantages and expertise of each grouping; NATO strikers, OSCE observers, …)

EU policy-makers remain confronted by a simple but unanswered equation: most see the institutional status quo as impossible; but, extra-EU closer cooperation is seen by many as divisive for the EU because of its lack of safeguards, while a proliferation of intra-EU ad hoc sub-systems is seen as even more deleterious. On the other hand, negotiations on substantive reforms (such as the adaptation of the acquis) as well as other procedural (the extension of QMV) and institutional reforms (the number of Commissioners) might well deliver inadequate solutions to meet the increased level of diversity and heterogeneity the next enlargement will introduce in the Union.⁽³⁵⁾ The need for more structured and integrated closer cooperation will then be impossible to ignore.⁽³⁶⁾ But, additional institutional innovations will then also be required because the limitations of the system of closer cooperation provided by Amsterdam are too severe. This seminar is therefore very timely and flexibility is in fact firmly on the agenda.

Some qualified doubts

A thousand years ago, during his reign in England, the Danish King Canute is said to have one day placed his throne in the midst of the waves ordering the tide to change. It did not. The Danes, who, taking their cue from Canute, tried to stop the process of European integration in June 1992, were more successful - but only marginally so. In fact, in large part it was the (first) Danish No to the Maastricht Treaty which prompted the debate in France and Germany that eventually led to the flexibility chapter in the Amsterdam Treaty.

Now, the first question one has to ask is whether the provisions for flexible (or enhanced) integration will be used at all. Three factors suggest that they may not:

1. the more hostile international economic and political environment with which the EU is now faced has led to a certain rapprochement between outsiders and insiders in the EU, not only in the monetary but also in the defence area.

2. the current ideological convergence in the EU around centre-left policies will tend to take some of the teeth out of the confrontationbetween insiders and outsiders in the EU, at least in the short and medium term.

3. as pointed out by Helen Wallace in her writings the divergence between policies in the applicant states and in EU member states are not as big as is often assumed, especially in pillars 2 and 3/title VI.

And yet, in the EU, arrangements set up to solve specific problems tend to take on a life of their own. Besides, flexibility also appears to have a more strategic dimension.

The essence of the flexibility provisions is legitimisation, not so much of variable geometry as of a Core-Europe model. If one looks carefully at the various provisions for enhanced cooperation one cannot but conclude that to all intents and purposes what is being set up by these provisions is a system of majority voting by stealth. A qualified majority of member states is normally able to authorize a smaller group of states to press ahead with integration - and those wanting to press ahead will often be the same qualified majority giving the authorization! As will be seen, flexibility fits in nicely in the neo-functionalist tradition.

In another sense as well, flexibility has a distinct neo-functionalist flavour, de-politicising what would normally be highly political, indeed constitutional, decisions: It can be seen as an elegant way for EU executives to bypass cumbersome national constitutional procedures, since in a number of policy areas enhanced cooperation now constitutes an alternative to further treaty changes. But the implications are ambiguous: While some integrationists may be tempted to try to deepen integration further without treaty changes and ratification, this pragmatism carries a (hidden) price tag. What will be the legal status of the acquis established within the framework of enhanced cooperation? And what will be the political legitimacy of such arrangements? In the Euro-sceptical member states governments under pressure from anti-EU opinion may use the provisions for flexible integration as an argument against new treaty reforms apart from the institutional reforms already scheduled. This points in the direction of a rather slim agenda at future IGCs. A final point. In some respects, the flexibility provisions are reminiscent of the old art. 235 (now 308), but unsurprisingly they also create some of the same problems as 235 - notably as regards democratic legitimacy. When shall we see the German Constitutional Court intervening against flexibility/enhanced cooperation ? You are probably all familiar with Fritz Scharpf`s concept of a "joint decision trap". Perhaps there is also such a thing as a separate decision trap.

In his excellent paper, Antonio Missiroli draws attention to the concepts of clubs and cores and quite rightly points to the lack of analyses that link these concepts to the study of the EU. Permit me to make a few remarks on the relevance of these concepts in our context: First, the theory of public goods assumes that free-riders cannot be made to pay. This is the argument about non-excludability. Yet, in the EU context, exclusion appears to be possible at least in the case of EMU. More generally, it would seem that one of the advantages of the EUcompared to unilateral forms of hegemony is precisely that, given the possibilities for creating linkages and playing nested games, free-riding in the EU is in actual fact rather difficult. One could also put it this way: the shadow of the present and future is rather long.

Secondly, in all clubs socialization is important. Effective socialization pre-supposes presence. Now, how can a club remain attractive, once it has had to exclude? Officially, closer circles of cooperation will be open to outsiders. Yet, not only is political practice likely to be slightly less open-minded, key provisions of the Amsterdam treaty are in fact distincly exclusionary:

a. outsiders opting into closer circles will have to accept an acquis established by others.

b. in some areas opt-out states, which later (due to a new domestic political climate) decide to opt in are almost treated as ordinary applicant states applying for membership.

All in all, while the possibility of enhanced cooperation may be a useful deterrent, it also has certain drawbacks. Apart from its legitimacy deficit, flexibility has a certain divisive potential. It must therefore be handled with care, otherwise it may end up defeating its purposes.

A comparison of national views

The „notion of flexibility" as a single unifying concept, is very difficult to establish in a satisfactory way, given the many different aspects it presents, according to the angle under which it is viewed and appreciated by analysts and member governments. Especially in EU integration, there are always the two basic approaches to view any accepted rule or legislation,

to view it as a given, a result of a past development and of interests acting, up to this point in time, and as a factor which will be applied, and create a certain effect, in the future,

to view it only as a snapshot of a occasional, one-off compromise/result between interests, forces, ideas, negotiated and fixed at a given point in time, in a constantly evolving integration-process -- in the teleological sense and in that of the concrete steps taken. For the last two decades, this process is increasingly marked by a sharpening contest between differentiation and unification as dominating principles in the further extension of European integration. 'Flexibility' as amended by Amsterdam is the principal and increasingly prominent form in which this contest expresses itself.

We concentrate on 'flexibility' as an object and result of a certain Intergouvernmental Conference, that of Amsterdam in 1997. Therefore it is very legitimate to treat flexibility mainly in the first of these two senses, and to subdivide and present it at the two corresponding levels of integration architecture, and of legal instruments of flexibility under Amsterdam. I myself have done so in writing and speaking on flexibility. At the same time I always feel that this frame of presentation may foster misunderstanding about the 'nature of the beast' itself, about the effects it might have, and the interests linked to it. Readers ask for instance: why are clear treaty clauses about proceeding on existing policies like EMU and Schengen, with opt-outs, treated under the same heading as enabling procedures under the treaty, opening the way to potential and as yet unknown future cooperation ventures?

To grasp this essential unity of the flexibility-phenomenon under its different forms/expressions, it has to be presented also under the second of the two above approaches (2.), the contest between differentiation and unification, constituted by:

· a continued push for more cooperation resp. integration, concerning discrete fields of government activity between certain member states (MS) of the EU, but not all of them,

· in which nonparticipating states resist successfully against being drawn into such initiatives, in unison

· with the tendency of the EU as a whole to recuperate and communautarize successful initiatives of this kind under its institutional frame and in its own procedures (as much motivated by the wish to harness the cooperation dynamic of some in the integration interest of all, as by that of restraining the advancing partners.)

In this contest, the results of Amsterdam concerning enhanced cooperation, represent:

· the differentiated legal form given to a status quo reached at this moment,

- when different initiatives of Enhanced Cooperation (EC), dating from different periods, had reached different stages of integration into the unified frame of EU institutions and procedures, whereas a (still unknown) potential of future Enhanced Cooperation had to be reckoned with,

- and when a specific balance between fully adherent and still abstaining MS existed (in spite of communautarization),

· and the concrete political and legal forms within which EU-MS may practice enhanced cooperation after Amsterdam, and until the next major change of the rules concerning EC between MS,

- by default, those areas where as yet unknown future enhanced cooperation still takes place outside procedures and institutions of the EU, in which EC remains 'free'

- those areas where known and proven past initiatives of external 'free' EC have been drawn in and established under the treaty frame (by consensus among the MS, but still without drawing in the hesitating ones), subjected to the common EC-rules or otherwise commonly established rules. Here, full validity of the new additions to the legal body of Union and Communities has to coexist with rightful nonparticipation rights for the abstaining members (PREDEFINED EC);

- those areas where the union institutions and procedures stretching out into as yet unknown future cooperation, are to constrain and recuperate such cooperation initiatives from the beginning (preserving the Acquis and member states' rights), at the same time as giving them the support of the Community apparatus (ENABLING EC) ;

- and finally those areas where the member states were not even able to go as far as this, even concerning old forms of 'free' cooperation, as in Common Foreign and Security Policy. Accordingly, there is only an easing of reaching foreign policy consensus in less contentious issue fields, in the pre-Amsterdam way, by introducing the possibility of 'constructive abstention' for some, making enhanced cooperation groups out of the others (CASE-BY-CASE EC).

1. Important differences exist between member countries as to the acceptability of flexibility in general, in its pre- or in its post-Amsterdam form, and as to the applicability of its different forms in the different sectors of the EU competence.

I will make a few brief introductory remarks as to these two aspects; and will conclude with

As to the acceptability in general, MS view flexibility in the context of two basic motivations, which every one of them shares at one moment or another, but in different combinations, making them into primarily progressive, or primarily status quo defender states:

- the motivation of those who want to advance in sectoral (or general) integration among member states;

- the motivation of those who want to preserve a given status of integration in which their interests are well taken care of, against the pressure of those who want to change it;

- (the day has already come when the right to preserve has also to be turned against those who want to retreat/renationalize. Agriculture being a case in point.)

The old treaty has always made the introduction of a new policy or competence under the treaty dependent on full consensus of all members, giving a structural advantage to the status quo defenders, over the progressive member states. This tended to push enhanced cooperation initiatives out of the treaty system proper. It is this imbalance, which has changed in recent years. Since Maastricht already, and concerning the introduction of integration deepening measures or of new EU policies, the legitimacy-premium of full consensus, compared to that of progress in integration has been increasingly reduced. In Amsterdam then, progressive member states saw themselves expressly, by the treaty itself, enabled to introduce new policies under the treaty system, without the agreement of all. This is the innovative amendment which Amsterdam has added to the unwritten and written rules which govern the utilization of 'flexibility' in European integration (predefined and case-by-case had already existed before).

And after all, the right of status quo defenders to preserve a given situation for themselves as well as they can, in the face of advancing partners, has been explicitly conceded as well, even buttressed with the emergency brake of new veto options in all three pillars.

Insofar, the Amsterdam amendment to flexibility is a treaty concession to the increasing tendency of differentiation in integration.

At the same time, and vis-a-vis this tendency, it reflects the continued effort of member states to uphold and strengthen the unity and comprehensiveness of the treaty system in the face of increasing diversity, by recuperating past grouped integration initiatives outside of the treaty system (begun in Maastricht with EMS, continued in Amsterdam with Schengen) and by encouraging future initiatives of the same kind to take root inside the system from the beginning.

Thus, if the premium on consensus and therefore the advantage for status quo MS has been reduced (concession to differentiation), the capacity of these last to control and constrain grouped advance initiatives introduced under the treaty has also been increased (concession to the unifying impetus). Integrationists and European policy strategiests (f.i. in Germany) feel encouraged in the outspoken hope that ongoing cooperation/integration interests of some can thus be harnessed to the aim of deepening integration for all.

It is within this frame of old motivations, of contending integration tendencies, and of a reformed legal setup that the general attitudes of member governments concerning the perspective of an EU with modified possibilities of differentiated integration in general, may be grouped and surveyed in some more detail. The legal setup of Amsterdam mirrors the contradictory aims behind it. Already today it fosters explicit demands to further reform on the occasion of the next IGC.

1. There is the majority group of countries which do in principle welcome further common extension of integrated policies. This does not mean that all of them support the Amsterdam amendments to the rules of flexibility. But only in this first group do they find full and outright supporters. This group includes all memberstates except UK, Denmark and Sweden.

2. There is the status-quo oriented minority group which rejects generalized further integration. In this one, on the contrary, these amendments were first opposed and only in the end reluctantly accepted. For this second group, the consensus-premium of the pre-Amsterdam treaty had been an important insurance that they could not be forced into a deepening dynamic inside the treaty, against their will. They are conscious of having lost a part of this insurance by the Amsterdam flexibility clauses. Here, we may group the UK, Denmark and Sweden.

We find three different types of attitude toward the new EC clauses, in which progressive or status quo orientation find distinct expression. Two of these concurred in supporting the EC-initiative in the run-up to Amsterdam:

- an interest of those who wanted to open the institutional frame for their enhanced cooperation initiatives and thereby to increase their integration potential for all,

- and an interest of those other supporters of this strategy who want to prevent future initiators of cooperation ventures from realizing them outside of the community, where other members cannot participate at will or influence them.

The first of these two attitudes seems typical for the two countries from which the idea originated in the first place, namely France and Germany, big member states having a lot of weight in EU decisionmaking, makers rather than takers of integration, confident that they would always participate in the important groups of enhanced cooperation and to have a decisive say on their creation. The second attitude, in comparison, seems to be a typical reaction of integration-minded member countries, which because of inferior weight and political means could not hope to trigger such initiatives themselves and --in the logic outlined above-- want to gain a voice in defining the kind of initiative they have to follow. Therefore they were and remain interested in the Amsterdam clauses which promise recuperation of free enhanced cooperation into the treaty framework, and that kind of institutionalized influence and assured access they look for.

As for the rest, both wanted to reduce the risk that such cooperation structures can become a challenge to the established Community.

But within this progressive group there is also a third attitude of those who hold a directly contrary view and see less the chance than the risk of letting enhanced cooperation enter into the treaty framework and be widely applied, not because they are opposed to further integration, but because they fear two negative effects:

- first, concerning cooperation initiatives that they approve: that - even if they want to - they will be not be able to participate in enhanced cooperation initiatives of other MS inside the institutional framework, because they lack the means to do so and cannot force the others to help them; (I happen to think that they can, but this is a different question);

- secondly, concerning cooperation initiatives that they do not approve: that the safeguards are not strong enough to prevent other MS from going ahead,and that the new rules might thus only result in opening up the EC system of institutions and procedures, made to serve all in the first place to serve just the ambitions and interests of a handful of them who would be able to press ahead with their particularistic interests, to leave the others behind, and thereby create a 2-tier-EU;

- this results in the frequent declaration that enhanced cooperation outside of the treaty system is considered less obnoxious than that inside in applying the Amsterdam clauses.

A widely-shared concern common to these last, but also among the principal opponents to EC, is the fear of an unbalancing of the traditional weights in negotiations, an unacceptable reduction of their bargaining clout in the EU where opposition to initiatives is often a means for states to get their favoured policies accepted in exchange for agreeing to the wishes of initiators, or the concern about being asent from decision (or opinion-) making circles, the decisions of which might impinge on their interests.

Let me make a few additional remarks which color and accentuate these general positions.

The first one would be that these were the basic positions at the time of the treaty negotiations, and they still seem to prevail when governments are asked for their general positions vis-a-vis enhanced cooperation of member states after Amsterdam. At the same time, and in spite of the differences, almost all auditioned governments (except Denmark, as far as I remember, but including the reserved ones) could imagine a more or less substantial amount of enhanced cooperation in certain specific fields of EU-competence. In fact, with few exceptions, it were only the technical ministries (namely for home affairs) which had at least a few clear ideas already about the uses to which they would specifically put the new Amsterdam flexibility clauses. General policy makers were much more difficult about this.

One more general point must be made: whereas at the time of negotiating the Amsterdam treaty the flexibility clauses stood very much in the foreground of the reform debate, about half of the auditioned governments feel now that they have moved away from the top of the agenda for the time being. The Dutch officials explained this in the most explicit way, by pointing to the pre-Amsterdam reasons for introducing the flexibility clauses:

1. Blocking role of the UK. After party-change and more pro-european attitude in UK gvt attention not to upset them by using EC against them;

2. EMU eventually not small and exclusive, but big and inclusive. Inclusive strategies have to be searched for;

3. Eastern enlargement will in any case give different accession speeds and modes. (Econ.Min.).

Characteristically, these were all governments which belonged to our category (1.2), positive integrationists but not enthusiastic about ECIF.

More than one government therefore expressed the expectation, that only after the coming Eastern enlargement of the EU the new EC clauses could move up on the agenda again and show their value. Here again there was not one single example given of how and where this might be applied.

Is there a general consequence concerning the favoured type of enhanced cooperation drawn out of this position after the treaty will enter into force ? In a very general way, the strongest defenders of the new clauses, in our groups (1.1) and (1.2), favour the Amsterdam amendment to enhanced cooperation over continuing the pre-Amsterdam mode of enhanced cooperation outside the treaty system, whereas those of (1.3) and of (2), less or more strongly reserved vis-a-vis the new possibilities, typically considered enhanced cooperation outside of the treaty system as less obnoxious than the Amsterdam mode within.

As to the general effects on integration deepening of amending the flexibility rules, there is a very forceful, but vague affirmation of pro-EC governments (1), in our hearings especially from Germany, that this would foster deepening - we have not heard one single concrete idea in this respect - whereas the contrary is affirmed among the group of pro-integrationists, reserved about EC (1.3). They openly express their concern or even suspicion, that the introduction of EC risked (or was even intended) to make future institutional reform and deepening - especially generalized QMV in Council voting - less urgent and necessary, and therefore reduced its chances of coming about.

The first pillar contains three very different policy fields which give rise to different positions on EC as well, (a) the internal market, (b) the economic and monetary union, and finally (c) the free movement of persons, especially within the Schengen application area, which is to be transferred from the 3rd to the 1st pillar.

a) Internal market: As to the internal market, the consensus between the member countries appears as the most complete. Every government stresses the tough conditions to be observed, and the sacrosanct quality of an open internal market, to be protected against any danger of new barriers via flexibility. Even so, some of them ventured out with small lists of topics where EC appeared feasible (or even desirable) under legal and economic considerations: in tax harmonization (D,NL,DK), or in services and intellectual property in international trade (I, not D,DK?), environmental matters (NL,D, not DK)

b) Economic and monetary union: Every Euro-government, even if it had been one of the doubtful candidates still recovering from the scars of this battle, placed priority on the good functioning of the Euro and would accept whatever EC appeared reasonable and necessary to this end, as to the decision-making procedure and the application of the necessary rules. There was striking contrast between relative equanimity of appreciation by Euro-countries, who are not discriminated against by the differentiated structure of the EMU-system, and the only auditioned member government of the opt-out group, namely Denmark, which carefully analyzed the resulting internal institutional and procedural differentiation inside EMU, and for whom the absence from important new decision-making bodies, including the ECB, proved a painful and clearly resented experience (even though it was the result of a Danish sovereign decision).

c) Free movement of persons, especially within the Schengen application area: Most auditioned governments (except Denmark, evidently) welcome the transfer of the Schengen acquis into the first pillar (via the third pillar, as the protocol to the Amsterdam treaty prescribes), some of them, for instance Spain, declaring to have wished a complete disappearance of the 3rd pillar and a transfer of its contents - with the addition of the Schengen acquis - into a new community policy for the first pillar. In Germany, too, wishes of this kind existed, but were more explicitly tempered by fears for the consequences which this was seen to bring to Germany in the field of immigration.

I already mentioned the remaining uncertainty about the legal possibility to utilize the new Amsterdam general enabling clause for enhanced cooperation in the institutional frame in the second pillar. The majority say 'NO'. There are some, f.i. in Germany, who would say YES.

Regret and dissatisfaction about the failure of the QMV-initiative remain in Germany, shared very explicitly by Italian officials and politicians, and possibly by others. But here again, very pragmatic voices have been heard from Denmark, and with the same tendency probably from the Netherlands and others: namely that QMV in the second pillar had never been tried since 1992, even where it was possible under the Maastricht treaty. One must therefore not place too many hopes in this possibility.

Nevertheless, there is this indirect new possibility of enhanced cooperation by default, introduced by the so socalled constructive abstention of the new article 23(1) of the EU treaty. When one or more member states utilized it in a decision on a common standpoint, strategy or action, adopted unanimously, this would make a group of enhanced cooperation in the institutional frame out of the remaining members.

Asked for their appreciation of this possibility, all governments seemed of one mind in declaring that constructive abstention could not be imagined for important foreign policy questions, where EU member state positions diverged and national interests of one or the other member state were involved.

For cases of lesser importance, positions were different. Again, Germany, true to its original intentions in this sector, sees the utility of constructive abstention, because a EU majority opinion on a foreign policy question of common interest is considered always preferable to the cacophony of fifteen individually nuanced national positions, even when one or several member states abstain. But it could turn out useful only for less important and for very specific issues. Positive example UN voting or UN and other international organizations/multilateral related cases. Of the auditioned governments, Spain shared this position, preferring a maximum of common actions supported by all governments, but agreeing to constructive abstention„ when a majority would be willing to act, so as to get ahead with CFSP." Other governments were clearly more reserved: the Dutch officials would prefer no EU position to one only presented by a majority, they saw the abstention of art 23(1) more as a destructive than as a constructive one, rather similar, it seems to me, to the Danes whom whe questioned on the same issue. In addition to this point of principle, both governments seemed to think that member states were rarely neutral enough on an issue of international politics to accept acquiescing in the art 23(1)-way.

As to security and defense policies, there was an interesting general similarity in a number of governments, especially, it seemed to me, in Southern Europe, where EC was not acceptable on the level of general integration policy, whereas the ministries of defense would see some interesting possibilities after all. Even considering free enhanced cooperation, opinions seemed to differ, concerning the range of issues still open for it. Starting with this, many officials felt that in security and defense most remaining fields were already regulated by WEU or by NATO rules; they saw little 'windows' for free EC: perhaps in arms production, trade and procurement. An especially outspoken priority was accorded to a functioning Atlantic Alliance (and its reform), by a number of governments: I remember Spain and Poland, two relative newcomers.

Even so, in Italy and Spain there was a certain regret that even military actions of a size which were accessible to WEU and did not require NATO, had to be put into place within a free cooperation/coalition outside of the WEU frame. Both defense ministries seemed to think that EC would be preferable because it would discipline and reshape security policy behaviour. Italian officials would like to include ad-hoc Petersberg-type missions like 'Alba', but also arms procurement.

Institutional Flexibility in a wider context

To begin with, the topic at hand is complicated by conceptual and terminological confusion, as, for example, Antonio Missiroli points out in his contribution to this volume. The term "flexibility" chosen in this article should be understood, on the one hand, as action taken by some but not all members of an international institution in an area within the competencies of the institution concerned. On the other hand, flexibility could also mean the participation of non-member countries in an activity of an international organization.⁽³⁸⁾

Activities which are covered by this definition may be further distinguished according to various criteria. Flexibility may be conceived as temporary in the sense that the other member states, provided they are able and willing to do so, are invited (and expected) to join later; or it could not be open to participation by all members of the institution. Flexible initiatives pertaining to the field of activity of an organization may be launched within it and on its behalf or outside and not under its aegis. The constituent instrument of an institution may already provide for flexibility; conversely, measures in which not all members take part might also be adopted without such explicit constitutional authorization.

Flexibility is particularly relevant for global institutions: not all problems they are faced with require solutions in which the entire membership is involved. In particular, regional issues may be dealt with successfully just by the members situated in the geographic area concerned. For instance, Art. 48 of the UN Charter introduces flexibility into the system of collective security of the United Nations. Under this provision, the Security Council may determine whether the action required to carry out its decisions for the maintenance of international peace and security shall be carried out by all the members of the organization or by some of them.

Moreover, the authors of the Charter devoted an entire Chapter (Chapter VIII) to regional peace and security mechanisms. Under Art. 52, para. 2, UN members shall make every effort at the pacific settlement of local disputes through regional arrangements or agencies to which they may belong before referring them to the Security Council. However, enforcement action under such arrangements or by such agencies may only be taken with the authorization of the Council. Regional institutions thus enjoy priority in the area of the peaceful settlement of international conflicts but are kept on a tight leash by the Security Council with regard to collective security action. Obviously, these Charter provisions are of particular relevance to Europe.

Flexibility was and is also practiced within the regional context of the EU, for example in the field of social policy, monetary union and under the Schengen Agreement. Against the backdrop of the obstruction policy pursued by the Major government in Great Britain and the prospect of the Union's next eastern and southern enlargement, flexibility was institutionalized in the Amsterdam Treaty in 1997 in considerable - perhaps too much - detail.

On the one hand, Title VII (Arts. 43-45) of the consolidated text of the TEU establishes a common legal framework for "closer cooperation" in the first and third pillars. Among the numerous requirements, safeguards and constraints laid down in Art. 43, para. 1, subpara. f is particularly far-reaching: Closer cooperation must not affect the competences, rights, obligations and interests of those member states which do not participate therein. This provision would seem to hardly leave any scope for flexibility, since it is difficult to imagine any initiative which would not "affect" the interests of the "outs" one way or another, either negatively or positively.⁽³⁹⁾

Further aspects of closer cooperation are regulated in Arts. 11 TEC and 40 TEU, respectively. The former provision adds more and more restrictive conditions for flexibility in the first pillar than in the latter for closer police and judicial cooperation in criminal matters.

The European Council further complicated matters at its Amsterdam meeting when it rejected, at the last moment, the proposed extension of the closer cooperation regime to the second pillar. Instead, the Union's political leaders introduced, in addition to "normal" abstention, the principle of "constructive" abstention for CFSP decisions at the Council level. If an abstaining Council member qualifies its abstention by making a formal declaration, it is not obliged to apply the decision but shall accept it as binding on the Union (Art. 23, para. 1, subpara. 2 TEU). Whereas it has to refrain from any action likely to conflict with or impede Union action based on that decision, it is not obliged to contribute to the financing of operations with military or defence implications it has not voted for (Art. 28, para. 3, subpara. 2 TEU).⁽⁴⁰⁾

The Maastricht Treaty also laid the foundations for internal flexibility within the EU in the area of defence. Whereas Art. 30, para. 6 subpara. (a) of the Single European Act had restricted security cooperation within the context of the EPC to the political and economic aspects, the CFSP deals with all questions related to the security of the Union under Art. J. 4, para. 1 TEU. In this context, the WEU is declared to be an integral part of the EU's development. The fifteen members of the EU have agreed to request the WEU composed of only ten of them to elaborate and carry out decisions and actions of the EU with defence implications.

The Amsterdam Treaty has added another facet to this division of labour by including the Petersberg tasks agreed on by the WEU in 1992 in Art. 17, para. 2 TEU. The WEU, with which closer institutional relations shall be fostered, is to provide the EU with the necessary operational capability. Moreover, when the EU avails itself of the WEU for Petersberg missions, all EU members are entitled to full participation; member states which contribute to the tasks in question take part fully and on an equal footing in planning and decision-taking in the WEU.

This new regime creates the "best of all (security) worlds" for the neutral and "non-aligned" members of the EU.⁽⁴¹⁾ It upgrades their previous observer status in the WEU to that of equal partners if they decide to participate in a Petersberg operation. At the same time, they do not have to abandon their neutrality or non-membership in military alliances.

The so-called Contact Group on Balkan issues, which is composed of four "big" EU members - France, Germany, Great Britain and Italy - and the United States and Russia, and the "G7/8", formed by these countries together with Canada, Japan and the President of the EU Commission, may be mentioned as important examples of flexibility outside the EU. These great power "directoires" are sometimes viewed with unease and suspicion by smaller European countries.

As regards the WEU, the Petersberg tasks are, on the one hand, an exercise in internal flexibility since participation by members in them is optional.⁽⁴²⁾ On the other hand, the alliance's ties to the EU constitute external flexibility because non-members which belong to the EU may decide to join its activities. Another "bridge" to the EU is built by the competence of the European Council to establish guidelines applicable to the WEU if the EU avails itself of the WEU - an interesting case of partial "Fremdbestimmung" from the WEU vantage point.⁽⁴³⁾ The external flexibility of the WEU is enhanced by the possibility that associate members and associate partners may also participate in its operations.

Similarly, NATO´s Partnership for Peace programme provides an another example of this type of flexibility, which, however, does not encompass "peacemaking" but is limited to cooperative missions based on the consent of all the parties involved.⁽⁴⁴⁾ IFOR/SFOR under NATO´s leadership may be mentioned as an impressive success story of an actual flexible operation to which many non-members of the Atlantic Alliance are still contributing. This mission has combined the agreement of all sides concerned with enforcement elements that go beyond traditional peacekeeping.

By contrast, the record of the EU and WEU in the field of crisis management is disappointing. In particular, the breakdown of law and order in Albania in 1997 would have lent itself par excellence to a Petersberg operation. However, disagreement on a WEU mission prevailed among member states, so that eventually an ad hoc mission led by Italy, "Operazione Alba", was launched in which WEU members and non-members took part in a "coalition of the willing".

Flexibility is a priori neither a positive nor a negative option for an international institution; its value depends on its actual use and the results achieved. If in a situation which requires a response by the organization not all its members - possibly together with non-members - take action on behalf of the organization and produce a satisfactory solution, flexibility is obviously to be welcomed. The organization's image and credibility will further be enhanced if it practices flexibility consistently whenever the need for it arises. Successful flexible initiatives necessitate the participation of those states which have the necessary resources at their disposal and whose interests are particularly affected.⁽⁴⁵⁾ By contrast, if a flexible reaction proves inadequate, it erodes the standing and legitimacy of the institution. The same is true if the necessary measures in one of its areas of activity are adopted by some of its members outside of its framework.

The desirable effectiveness of flexibility under the aegis of an organization can certainly be improved if organizational and other preparations are made in advance. Such steps are particularly important in the area of security where operations cannot be launched from scratch overnight.

Pros and Cons of Ad-hocery

Life, and especially diplomatic life, knows a great deal of ad-hoc decisions, reacting to unexpected developments in the environment in which we have to function. Most of what we are doing is ad-hoc and it is a great challenge to maintain a line of consistency in our behaviour.Diplomacy tries to find support for specific positions which inherently includes an element of changing coalitions,varying from issue to issue. Since World War II the process of European integration and the creation of a transatlantic alliance have introduced institutions which put policy formation into a permanent framework. Without institutions nothing would last. Therefore, the notion of ad-hocery is characterised by action outside institutional procedures to deal with a specific problem. In the field of security we know the coalitions of the able and willing to engage in an operation, which proved to be a necessary complement to action by NATO or WEU, but lacked the long term planning on which collective defence was based. Without pre-planning quick reaction is not possible. Especially, combined arms cooperation involving major inter-service cooperations are difficult to imagine in an ad-hoc context, unless they are centered around an existing US command.

Before the European Council meeting in Amsterdam the concept of flexibility was very topical. Since then interest has abated, for two reasons. The first one was the changed atmosphere in London, where the Blair government was adopting a constructive attitude towards Europe even though in the short term British policy would not undergo great changes on substantial issues. For the French and Germans flexibility had been a remedy of last resort if the UK were to continue to be negative on closer union. The second reason was the feasibility of starting monetary union with a much larger group than originally expected. If EMU had to be restricted to a small number of countries, concentrating on Germany, France and the Benelux, it would have served as the hard core of enhanced cooperation . Only in the field of military operations the British cooperation would be sorely missed. But with a monetary union of 11 member states and dispersed attitudes on other issues, there was no incentive for pressing ahead with a core group.

Moreover, Amsterdam had made flexibility very difficult to implement. Understandibly, the priority of keeping the common market from desintegrating was a powerful argument against limited groups making special arrangements. The third pillar and the Schengen area remained intergovernmental, at least for the near future, with little impediments for special arrangements like the Benelux. In the second pillar enhanced cooperation was specifically excluded by the introduction of the notion of "constructive abstention" and the possibility of preventing qualified majority voting for reasons of vital national interest. This recurrence of the so-called Luxembourg compromise - which obliged the country involving it to explain the reasons - was also included for the third pillar. It did not provide a watertight solution for this contingency, because it provided only for reference of the issue to the European Council, which would deal with it under the procedure of unanimity.

Flexibility or enhanced /reinforced cooperation is a form of variable geometry as distinct from multiple speed arrangements. Under the latter the ultimate objective is identical for all, but some need more time in reaching it. Transitional provisions in the accession treaties are examples. Variable geometry on the other hand does not exclude the possibility of other countries joining, but applies basically to a limited group of states engaged in a specific field of cooperation. WEU, EMU and Schengen belong to this category. Neither term relates to ad-hocery, which denotes joint action on a specific issue or in a specific operation.

At the NATO Summit of January 1994 the principle was agreed of making NATO assets available to WEU or to an ad-hoc coalition. Attention was focussed on WEU, but action by a group of countries coming together for a specific purpose was not excluded. After all, Desert Storm had been an ad-hoc coalition to expel Saddam Hussein from Kuwait, created under US leadership with some 29 countries joining, but not in the framework of the Alliance. With the expansion of the Partnership for Peace program of NATO and the admission of Associate Partners of WEU both organisations seemed to become sufficiently flexible to gear their operations to the number of countries participating in the operation, which was prompted earlier by the German difficulty in sending military personnel outside the NATO area. The mine clearing operation in the Gulf in 1987, the embargo against Iraq in 1990, the arms embargo and subsequent blockade in the Adriatic and on the Danube in 1992 and the police action in Mostar never included all WEU members. Originally NATO was constrained from out-of-area operations which gave WEU a niche of its own, but with Sharp Guard declared the out-of-area debate out of date. In doing so, NATO also had to adapt its force structure to the implications of operations shifting from collective defence to peace support and intervention. Participation by all members could no longer be assumed, not even in the regional commands, as individual assessment in capitals was likely to vary. The emphasis on the Rapid Reaction Corps, consisting of a catalogue of forces, later followed by the concept of Combined Joint Task Forces - enhancing NATO's internal flexibility as well as its availability for operations not led by its command structure - were a suitable response to new conditions.

Both NATO and WEU assumed that the political decision to commence an operation required the consensus of all full members. Flexibility would only exist in its implementation. With an expanding membership the unanimity rule might have to be relaxed. The NATO treaty itself contains only one provision requiring unanimity, i.e. for the admission of new members. In practice, some flexibility already has crept in, first by the non-participation of France in the integrated military command structure, later by repeated instances of footnotes to Alliance communiqué's, distancing a member state from specific paragraphs.

Not for the first time theory and practice did not coincide. In the Albanian crisis of the spring 1997 neither NATO nor WEU were prepared to act and Italy took the lead in creating an ad-hoc coalition for operation Alba, which included among others a WEU observer, Denmark, and an associate partner, Romania. The organisation took a leaf out of the WEU book, by forming a political body in Rome, resembling the WEU Council, but chaired by the Italian political director. Full marks for Italy, but a missed opportunity for WEU, which was too easily diverted from taking action by spontaneous negative reactions from Germany and the UK at an informal CFSP ministerial at Apeldoorn in the Netherlands.

Recent planning for the extraction force for Kosovo, to be stationed across the border in Macedonia, seemed to break new ground in the relationship between the organisations involved in European security. OSCE is organising a 2000-men strong mission of unarmed "verifiers", NATO is monitoring from the air, the extraction force comes under French command as the "framework nation", but functions in the NATO line of command. The head of the OSCE mission will have a direct line with SACEUR to request activation of the extraction force whenever his people are in danger. At the time of writing the precise arrangements are still unclear, including the reliance on the promise President Milosevic that he will guarantee the safety of the verifiers. At first glance, however, doubts remain whether the lessons from Bosnia have been heeded: a clear mandate, unambiguous political authority and unity of command. We hope for the best. It seems clear, however, that the EU/CFSP as well as WEU have taken themselves out of the loop even more than in Bosnia, leaving the operation largely to the US and taking dispersed positions on the Contact Group. Paradoxically, the fact that the extraction force does not contain US personnel, except perhaps in the headquarters, puts paid to the assertion that Europe would not be able to act alone in these situations. Why then not call it a CJTF if this concept still has any meaning?

One way of avoiding centrifugal tendencies in the future would be to emphasise the role of multinational forces for crisis management at all levels. For collective defence multinationality below the corps level was considered ineffective, for peace support missions it is essential to demonstrate sollidarity, to avoid disproportionate risk for a single country and to gather sufficient forces. In the new missions the difference between larger and smaller countries disappears as many contributions are of batallion-size or even smaller.

Of course multinational forces have nothing to do with ad-hocery. On the contrary, they introduce a new element of commitment in defence planning which may be sorely needed to avoid a slippery slope in declining defence budgets. They should involve training and exercise for peace support operations and should be seen to be used in these missions. To allow for early use contingency plans should be developed including force packages and arrangements for command, communications, transport and logistics. During my period as WEU Secretary-General it was not possible to engage in such contingency planning, largely because of German reticence out of fear of political sensitivity in the Bundestag. Times may have changed with German participation in SFOR, thought impossible four years earlier, and willingness to join the extraction force for Kosovo. Likewise, pre-planned NATO or WEU units could be offered to the UN or OSCE, without waiting for a request (which would probably not be forthcoming anyway).

On the political side the picture is blurred. If institutions are ineffective in crisis management, they will be by-passed by ad-hoc arrangements. The Contact-Group for Bosnia was such an expedient (which by its extension to KOsovo is assuming permanent proportions). The new fashion of forum-hopping from one international organisation to another acquired an absurd character by ministers speaking with different tongues on each occasion. The notion of interlocking institutions sounded wonderful, but has not been substantiated in practice. In this context the CFSP should concentrate more on specific policies, rather than on the unwieldy objective of an all-encompassing foreign policy. The declaratory character of most of this position has undermined its credibility. The appointment of special representatives for specific missions should be encouraged as a more practical arrangement than the future High Representative, certainly as long as the complicated arrangements for common strategies have not proved their value and the policy-initiating role of the new dignitary have not been established.

This paper did not deal with flexibility on the equipment cooperation side, where the systematic approach continues to battle with project orientation. Conceptually the two are not mutually exclusive, but the field is littered with legalistic objections which probably hide unwillingness to make real progress. As a result industrial consolidation takes place without governments - the only buyers - agreeing their priorities and channelling their research money accordingly.

"Ad-hocery" is a neutral concept

"Ad-hocery" is not a concept which is inherently good or bad. Its desirability or otherwise depends entirely on the view which is taken with regard to two basic sets of assumptions governing the debate on the Common Foreign and Security Policy (CFSP) - those concerning the preferred nature of the CFSP, and those concerning the relationship between the intergovernmental and the integrationist approaches. The side of the fence you come down on on these issues will determine your view on whether "ad-hocery" is a good thing or not.

The first set of assumptions concerns the desirability or otherwise of the European Union (EU) having the capability of projecting power. Should the EU be able to implement its policies including through the use of force, in imitation of the classical state, or should it rely on persuasion and non-violent coercion? To put the question in a misleadingly crude fashion, should the EU be a "military" or a "civilian" power? The tendency in recent years, dating from the Gulf War but gathering momentum because of frustration over former Yugoslavia, has been in favour of the former, although reluctance to take concrete steps suggests that thinking is still influenced by the latter.

To adjudicate between the two is not here the issue. The point is that, if the EU is to project military force, it would clearly be irresponsible as well as ineffective to leave the modalities to inspired improvisation. Lives are at stake: they cannot be exposed to unnecessary risk by inadequate forward planning, which presupposes the existence of stable structures. Furthermore, the credibility of military intervention depends to a considerable extent on expectations, which by definition are low if action depends on ad hoc solutions. The strong emphasis put on analysis and planning in recent European security debates, and the creation of the WEU planning capability, is therefore fully justified.

If, on the other hand, the EU is to act as a civilian power, improvisation is less damaging. It would be going too far to claim that it is preferable in all circumstances to planned action, but the advantages of flexibility and creativity it frequently brings to the foreign-policy making process can be had without paying a price which a "military" power would find too heavy. This is particularly beneficial when an infant foreign-policy actor like the EU is still finding its way. The first few months of the EU's activity in former Yugoslavia were a good example of the benefits of improvisation. Conversely, the involvement of UNPROFOR demonstrated the dangers of improvisation for military intervention.

The second set of assumptions, concerning the respective merits of the intergovernmental and integrationist approaches to European foreign-policy making, leads us in a slightly different direction. It is commonly agreed that European Political Cooperation, which not only operated on intergovernmental principles but was set up in order to preserve those principles in the foreign policy field, succeeded in making regular progress throughout its two decades of existence by improvising as required and then codifying the results of improvisation. The Copenhagen and London Reports, the Stuttgart Solemn Declaration, and the Single European Act are all examples of this, as are the innumerable minor details of practice and procedure enshrined in the ci-devant Coutumier. To use the same example of former Yugoslavia, EPC in the first months came up with important procedural innovations - Troika diplomacy, the EC Monitoring Mission, the Carrington/Owen Conferences - which lacked only the benison of success to take EPC into a significantly strengthened new stage of its development.

As it was, the pressures exerted by the changes in Central and Eastern Europe and especially the unification of Germany took the EU alone a different path, that of the CFSP, which was an attempt to reconcile integrationist ambitions with intergovernmentalist reality. The most significant integrationist contribution to EU foreign-policy making may well prove to be, not majority voting and the new role of the Council and Commission in the CFSP, but the creation of a legal framework setting the limits within which the CFSP may develop. While not totally excluding improvisation, it makes it very difficult, especially as the lawyers have taken over from the diplomats the responsibility of deciding when improvisation is appropriate. There was some leeway left for ad-hocery in the form of partial participation (i.e. by some but not all Member States) in the CFSP-type actions, which while dubious on political grounds had some practical advantages, but this too has now been regulated by the Amsterdam Treaty. Improvisation has been legislated out of existence by the integrationists; given their beliefs, they had no choice.

- military integrationists can be consistent as regards ad hoc solutions - they must be against them;

- civilian intergovernmentalists can also be consistent - they must favour ad hoc solutions;

- military intergovernmentalists and civilian integrationists must be schizophrenic, now favouring ad hoc solutions, now opposing them.

Of course, real life is not like that. But the argument illustrates a fundamental truth about what is currently stultifying the CFSP: that the more serious it gets, especially over the use of force, the more it needs to operate within solid legal structures, and the more it operates within solid legal structures, the less likely it is to be able to react flexibly to unforeseen challenges.

Looking to the future

For adjusting expectations, let me begin by saying that, in a lion's cage, the menu is not served the French way - nicely decorated, in a carefully calculated order - but it is rather the raw meat that has been thrown in, you do not want to get too close. Let me suggest to you, in the way of being in a lion's cage, to just throw a few things at you and see what we get out of it.

First of all, I would like to begin with two defining observations. The first one relates to our previous debate. My sense is that, in the present situation, and given the inward focus of democracies (which I do not criticise because I believe it is natural - that is after all what we always wanted: not to be forced to care too much about what happens elsewhere), we are basically all in an opt-out preference situation. Foreign policy is about controlling externalities, is about containing certain risks and, except for a few who pursue wider goals, we are all very much satisfied if we are not asked to actually go somewhere, shape events somewhere, commit resources somewhere. And those who share those wider goals, oftentimes have to defend themselves against the accusation that they do so for the general purposes of their own national greatness. This being so, others are on alert not to let these actors go too far and frustrate their national ambitions.

Second defining remark: today and yesterday, we have talked about flexibility in a very wide sense and I would, rather than talking about the multiplicity of potential opt-outs and stay-aways, rather than dealing with doubts and desires, talk about the other sense of flexibility, which is enhanced cooperation as an opt-in strategy. In my view, flexibility is a strategy to realise some ambitious goals than to deal with the fact that heterogeneity is growing among us. In this sense, my general understanding of enhanced cooperation is that it is an instrument to realise treaty goals or necessary common goods, even if a number of member states is unwilling or unable to do this because the realisation of these goals - or the delivery of these goods - is in the general interest of the Union at large.

Now, what does Amsterdam have to say about flexibility and enhanced cooperation in this direction? First of all, it offers what we have been discussing, constructive abstention. Whether it is constructive I think we will have to see. By looking at the way it is set up, I have some doubts about that because constructive abstention is almost consensus or "consensus minus one" but not the one who cares about a certain issue but minus the one or very few who do not care. So I would say that the many other references we would have in international organisations to the consensus-minus-1 principle is not the understanding of constructive abstention in the Treaty. Therefore, it is not Greece being interested specifically in the State of Macedonia which is a reference point for constructive abstention. Rather it is consensus with some free-riding possibilities.

There are two other ways in which the Treaty allows for some flexibility. The first is of course the already-mentioned common strategy / joint action package, the majority voting option in the action phase. And here I believe that the question marks and reservations that have been raised in this debate - about what it would do to the ability of the Union to actually come up with a set of strategy decisions to build on them later - is very valid. The second option in principle is the notion to mandate WEU to act. This option, too, has a drawback: on the one hand, I think it is very appropriate (and Wim has made that point) by looking not legalistically but pragmatically at the varied membership and qualities inside WEU, the drawback is the same principle, turned around. With the Amsterdam Treaty, non-committed members of the European Union (that is non-full members of Western European Union) actually have the possibility to decide not only on the mandate to WEU but also to decide and to take part in the follow-up decisions of WEU, which is good if you have a group of countries which actually want to use WEU, but which is bad when we have a pattern - which we also discussed yesterday and today - where groups of states, or certain member states, may have an interest in using this opportunity to make sure that the implementation of such a mandate does not go an inch beyond what they would like WEU/EU to do. So, in general, we cannot really rely on the option that, on the basis of a strategic decision, all further implementation issues will be given to WEU, and that WEU - at least among its full members - would then have the opportunity to interpret this strategic mandate and to act accordingly. Thus this specific stipulation of the Treaty which allows for non-WEU members to be involved in the follow-up decisions inside WEU will function as a safety valve.

Let me try to approach the issue from another angle by looking at the question of what we need in relation to a potential enhanced cooperation inside CFSP. Let me give you a 3 +1 answer. I do not want to argue the thesis I will present but rather leave that up to the debate.

First of all, Europe needs a stick to negotiate a peace with those who do not depend on or believe in the logic of the carrots. I think that this is something that -without reference to the application of civilian means, we have to bear in mind the ineffectiveness and the poor cost-efficiency of our civilian crisis and conflict management - has to do with exactly this issue.

Secondly, Europe needs visible security providers to maintain the credibility of the complementarity between European Union and WEU on the one hand, and NATO on the other hand. We have mentioned in this morning's debate that there is now wide complementarity in the membership of EU and WEU, bearing in mind the various forums; but the essential point comes with the second stage of enlargement, when we actually move into a situation in which potential EU members may not be allowed to join NATO. Whereas I think that it is very clear that among the present 15 members: no single member could be denied NATO membership if that country would wish so. This principle is very important for the European Union, maybe more important than for NATO. If we do not want to lose this, we also have to reflect on what we would need to do in which form to convince at a given point the non-EU NATO members (including the United States, including Turkey or Norway) that it is justified for them to take on this extra commitment in the Baltics or in South-Eastern Europe.

Third, Europe needs more efficient procurement and arms production: for the defence of a much-enlarged Union, for the defence of Europe's defence industrial base and as a technology resource for smaller armies which we will have in the future, with rapid action and reaction and low-casualty risk capabilities. At present, we have exhausted the mechanisms at hand here. Juste retour, I believe, is a joke. The efficiency of joint programmes is questionable. Wim was very right to say that it is either too early for this traditional set of multinational cooperation or it is too late. Mostly, when it is too late, it also means that we are paying too high a price for that. And I do not see how we can make progress under the present structures, but we might with and through enhanced cooperation.

The Plus-1 argument would be that we need political will. The present strategy of building CFSP from EPC to the Amsterdam Treaty has been to build a house and to enlarge it, expand it, make it more comfortable and to wait for the moment in which the political will arrives to take residence in that house. The latest addition to this building may finally attract the sufficient political will, but we cannot be sure about that. I do not want to speculate about how things look like once the planning unit exists and once issues are discussed based on just one dossier, having the European viewpoint as the starting point of the foreign and security debate inside the Union. My guess is that this is a chance but not a promise. The political will needs a focal point, a point of crystallisation, as I understand Tony Blair has called it the other day, a core of consensus and shared responsibility. So my suggestion would be that, in order to have a political will to become attracted to the house called CFSP, it may need an engine which, by its existence, forces other member states to think of whether or not they should become more involved in this process. What does it mean for enhanced cooperation? I think the flexibility acquis in this sense in the Amsterdam Treaty does not sufficiently respond to the needs as I have tried to describe them. It rules out, as we have discussed, coalitions of the willing, and thus will not help to develop CFSP into a mandating institution. WEU in my reading - in some disagreement with what Wim has said - is already too large and it is too diverse to become such a coalition that would combine the elements of ad-hocery on the one hand and preplanning and a preconceived action capability on the other. So it will neither be able to serve as a coalition nor will it be an engine to CFSP. The current Treaty does not build, against everything that has been said about large and small, on the law of gravity, which after all applies in international relations and should be borne in mind. So CFSP, in its present structure including Amsterdam, is still no place for power actors inside the EU or for the EU members inside the Contact Group. They do not get a privileged place there to express their commitment visibly enough (also to their own national constituencies) to justify added commitments and resources in this process.

Security and defence integration, which are the steps 2 and 3 in the cascade of both the Maastricht and the Amsterdam rhetoric, is unlikely to materialise in such an environment, although the strategic situation, the budget constraints, the defence technological developments speak in favour of such a development.

My conclusion is that enhanced cooperation will have to seek ways outside the Treaty, but in the light of the Treaty, and outside the EU, but on the basis of the performance abilities of the Union. An initiative on creating an engine for CFSP would thus start on the basis of bi- or multi-lateral cooperation and that is exactly how I understand the hints the British Government was giving the other day, with all of the implications to stir up the debate inside the Union as well as to stir up responses from principle partners.

Against this background, I see two options for the future, or rather two starting points for enhanced cooperation.

The first option would be to develop CFSP more clearly and visibly, into a mandating institution. An instrument of that would be the Common Strategy. Joint actions could then be based, not only on QMV but also on flexibility. If one wishes ad hoc coalitions acting in the name of the Union very soon adding up a history of ad-hocery forces, the question will arise whether or not there should not be a certain form of preplanning. And those who will always be present in those ad hoc operations will be the first to raise that question - partly because it is a way for them to assert their relative position among the EU member states. Certainly, this means that these coalitions will be using their own means but they will act - and there the Treaty offers some room for cofinancing or for basic solidarity demonstrations on the part of the other member states.

The second option would be to create an engine by using the EMU experience, that is a core emerging in CFSP, or alongside CFSP. It would probably not build on foreign policy or security policy, but on common defence plus security policy, and it would offer the prospect of a joint defence structure. A smaller group could, if it only wanted to, fairly soon establish an internal market for arms products, for military industry among themselves. The pressure on governments in this direction is growing. Everyone who has followed the Aérospatiale-British Aerospace-DASA quarrels and changes over the past year knows that we are moving into a decisive phase where either British Aerospace will break out or there will be a new scheme for cooperation. In such a situation, actually, such a core might think of the possibility of not only integrating certain capabilities in terms of rapid reaction but also in terms of territorial defence. And if we look at it - it seems to me that pooling resources in territorial defence has never been as easy to do as under the strategic situation of today - this combination of establishing backbones for rapid reaction, of building a joint arms procurement market and of trying to apply efficiency to territorial defence could be a sufficient engine to make everybody else inside CFSP think of what their own position would be like and whether or not they would not be interested in joining at one point. I would presume that the clout that such a group would bring into CFSP policy bargaining could be a sufficient added weight to the vote of these countries to speed up the debate.

Concluding Remarks

Let me briefly summarize the main points, or rather issues raised during our discussion. One of the main difficulties we have come across has certainly been terminology, namely the translation of political slogans and policy preferences into clear concepts and definitions, be it for 'flexibility' as such, or for enhanced/reinforced cooperation, or for 'ad-hocery'. It is a fact, however, that now we have to deal with how to achieve real "closer cooperation" as enshrined in the Amsterdam Treaty. A first issue we have addressed, albeit without coming to a final assessment, is whether those provisions will ever be put into place: in other words, will they actually be enacted - and, if yes, how and by whom - or will they mainly serve as a sort of institutional deterrent against political boycott and "forum hopping" ?

A second issue we have singled out is the distinction between institutional flexibility and 'ad-hocery': if it is true, in other words, that a certain measure of 'ad-hocery' is inherent to foreign and security policy action per se - given the specific configuration of each contingency and the fact that military operations, in particular, cannot be evaluated only on their institutional implications and gains - one thing is to practice forms of flexibility inside, and quite another to achieve it outside of existing institutions and organizations. At the same time, it may well happen that, in order to overcome the present stalemate and set positive precedents, we will try and put some 'European label' on actions not undertaken by all EU members, thus blurring a basic distinction. In this respect, the question was also raised whether 'Operation Alba' should be considered a sort of unicum , a one-off contingency - given the US reluctance to be involved and the Germano-British reluctance to use existing European institutions - or rather a test-case for possible future crises: the answer will also determine whether or not 'Alba' should be considered a missed opportunity for using the European 'label'.

Finally, if differentiation is to lead to the formation of 'clubs' inside Europe, it is worth wondering whether there will be 'subscribers' to all of these clubs, capable of forming visible 'cores' across different institutions and policies. How sizeable could/should such 'cores' be ? Should they be restricted to the 'Big Three' - or four, or even more - or should they try and keep the traditional European balance between big and small countries ? On the one hand, a very limited directoire could be more acceptable e.g. for small and medium-sized countries than a slightly larger but still exclusive one. On the other hand, if permanently excluded from the preliminary decision-making process, the same countries could be tempted to search for an even bigger hegemon, i.e. the United States. Paradoxically, however, a fundamental change of attitude on the part of the US - either way, more pro-European or instead more unilateralist/isolationist - could be crucial to helping strengthen CFSP. And, needless to say, the 'dual' enlargement of the EU and NATO will generate further differentiation and will demand more 'flexible' arrangements.

All these questions - and presumably other ones - are bound to remain on the table. They will keep concentrating our minds and stimulating our discussion. We thank you all for your valuable contribution and look forward to the next stages of a debate that will undoubtedly pick up speed in the very near future.

About the authors ....

Christian Deubner is the Head of the Research Group on Integration and External Relations in Europe at the Stiftung Wissenschaft und Politik in Ebenhausen, Germany

Willem van Eekelen is a Senator in the Dutch Parliament in The Hague. He was Secretary-General of Western European Union from 1990 t o1994.

Josef Janning is Deputy Director of the Centre for Applied Policy Studies of the University of Munich, Germany

Antonio Missiroli is a Research Fellow at the WEU Institute for Security Studies

Hanspeter Neuhold is a Professor at the Institute of International Law and International Relations in Vienna, Austria

Simon Nuttall is a Visiting Professor at the College of Europe in Bruges, Belgium

Thomas Pedersen is Associate Professor in the Department of Political Science of Aarhus University in Denmark

Eric Philippart is Associate Director of the Centre Facultaire de Recherches Internationales of the ULB in Brussels, Belgium

1. ¹ See Christopher Hill, The Capability-Expectations Gap - or Conceptualising Europe's International Role, Journal of Common Market Studies , September 1993, pp.305-328. For an overview cf. Jolyon Howorth, Towards a European Foreign and Security Policy?, in Jack Hayward, Edward C. Page (eds.), Governing the New Europe (Cambridge: Polity Press, 1995), pp.317-345; Anthony Forster, William Wallace, Common Foreign and Security Policy', in William Wallace, Helen Wallace (eds.), Policy-Making in the European Union (Oxford: Oxford University Press, 1996), pp.411-435; Michael Smith, The EU as an International Actor, in Jeremy Richardson (ed.), European Union: Power and Policy-Making (London: Routledge, 1996), pp.247-262; Stelios Stavridis, The Common Foreign and Security Policy of the European Union: Why Institutional Arrangements Are Not Enough, in Howard Machin et al. (eds.), New Challenges to the European Union: Policies and Policy-Making (Aldershot: Dartmouth, 1997), pp.87-122; Spyros A. Pappas, Sophie Vanhoonacker (eds.), The European Union's Common and Foreign Security Policy: The Challenges of the Future (Maastricht: EIPA, 1996); Martin Holland (ed.), Common Foreign and Security Policy: The Record and Reforms (London: Pinter, 1997); Peter van Ham, La construction d'une Europe politique: la PESC, in Anne-Marie Le Gloannec (ed.), Entre Union et Nations. L'Etat en Europe (Paris: Presses de Sciences Po, 1998), pp.227-265; Marie-Francoise Durand, Alvaro de Vasconcelos (eds.), La PESC. Ouvrir l'Europe au monde (Paris: Presses de Sciences Po, 1998); Jan Zielonka (ed.), Paradoxes of European Foreign Policy (The Hague: Kluwer, 1998); Kjell A.Eliassen (ed.), Foreign and Security Policy in the European Union (London: Sage, 1998); and John Peterson and Helene Sjursen (eds.), A Common Foreign Policy for Europe ? Competing Visions of the CFSP (London-New York: Routledge, 1998).

2. ² See e.g. Patrice van Ackere, L'Union de l'Europe Occidentale (Paris: Presses Universitaires de France, 1995); Eric Remacle (ed.), La PESC, l'UEO et la Cig (Bruxelles: Editions de l'Institut d'Études Européennes, Université Libre, 1996); Philip H. Gordon, 'Does the WEU Have a Role?, The Washington Quarterly, Winter 1997, pp.125-140; Robert Antretter (ed.), Sicherheit in Europa - Die Westeuropäische Union, transnational, 35 (Bonn: Europa Union Verlag, 1997); Anne Deighton (ed.), Western European Union 1954-1997: Defence, Security, Integration (Oxford: St.Antony's College, 1997), pp.93 ff.; G.Wyn Reese, The Western European Union at the Crossroads: Between Transatlantic Solidarity and European Integration (Boulder: Westview Press, 1998); Willem van Eekelen, Debating European Security 1948-1998 (The Hague: SDU Publishers, 1998), particularly valuable as the author served also as Secretary General of WEU from 1989-1994; and Andre Dumoulin, Eric Remacle, L'Union de l'Europe Occidentale - Phenix de la defense europeenne (Bruxelles: Bruylant, 1998).

3. ³ See for instance Gunnar Sjöstedt, The External Role of the European Community (Farnborough: Saxon House, 1977); Geoffrey Edwards, Elfriede Regelsberger (eds.), Europe's Global Links: The European Community and Inter-Regional Cooperation (London: Pinter, 1990); David Buchan, Europe: The Strange Superpower (Aldershot: Dartmouth, 1993); Ole Norgaard et al. (eds.), The European Community in World Politics (London: Pinter, 1993); and Louis Balmond, Jacques Bourrinet, Les relations extérieures de l'Union Européenne (Paris: Presses Universitaires de France, 1995). For a thorough discussion of the (controversial) concept of "civilian power" as applied to the EC/EU - it was first coined in 1973 by Francois Duchêne (see also below, fn.21) - cf. Hedley Bull, Civilian Power Europe: A Contradiction in Terms?, Journal of Common Market Studies , 1-2, XXI (1982), pp. 149-170. For a recent reassessment in the light of the forthcoming EMU and its interplay with the EU's external relations and CFSP see Carl Bildt et al., What Global Role for the EU? (Bruxelles: Philip Morris Institute, 1997) as well as Richard G. Whitman, From Civilian Power to Superpower ? The International Identity of the European Union (London: Macmillan, 1998); Carolyn Rhodes (ed.), The European Union in the World Community (Boulder: Lynne Rienner, 1998); Christopher Piening, Global Europe: The European Union in World Affairs (Boulder: Lynne Rienner, 1998); Carolyn Rhodes (ed.), The European Union in the World Community (Boulder-London: Lynne Rienner, 1998). Unfortunately, the institutional competence over 'external relations' is presently shared by at least four different Commissioners, thus hampering policy consistency and weakening its overall impact and visibility. Such fragmentation may be overcome by regrouping the different regional and functional portfolios under the responsibility of a single Commissioner, who would also act as Deputy President. Proposals in this direction are being considered for the next Commission, set to take over in January 2000.

4. ⁴ For the latest discussion see e.g. Graham Avery, Fraser Cameron, The Enlargement of the European Union (Sheffield: Sheffield Academic Press, 1998); John Redmond, Glenda G. Rosenthal (eds.), The Expanding European Union: Past, Present, Future (Boulder: Lynne Rienner, 1998); Francoise de la Serre, Christian Lequesne, Vers l'elargissement de l'Union: integration ou implosion ?, in Francoise de la Serre, Christian Lequesne (dir.), Quelle Union pour quelle Europe ? L'apres-traite' d'Amsterdam (Bruxelles: Editions Complexes, 1998), pp.125-156; and Pierre-Henri Laurent, Marc Maresceau (eds.), The State of the European Union , vol.IV, Deepening and Widening (Boulder: Lynne Rienner, 1998).

5. ⁵ That is, the European Political Cooperation framework established at the EC intergovernmental level in the early 1970s. Unquestionably, 'the' book on EPC is Simon Nuttall, European Political Cooperation (Oxford: Clarendon Press, 1992). See also David Allen, Reinhard Rummel, Wolfgang Wessels (eds.), European Political Cooperation (London: Butterworths, 1982); Philippe de Schoutheete, La Coopération Politique Européenne (Bruxelles: Labor, 1986); Alfred Pijpers, The Vicissitudes of European Political Cooperation: Towards a Realist Interpretation of the EC's Collective Diplomacy (Leiden: Leiden University Press, 1990); Martin Holland (ed.), The Future of European Political Cooperation: Essays in Theory and Practice (London: Macmillan, 1991); and Elfriede Regelsberger, 'European Political Cooperation', in Jonathan Story (ed.), The New Europe: Politics, Government and Economy since 1945 (Oxford, Blackwell, 1993), pp. 324-336.

6. ⁶ For a detailed overview of the early stage of the discussion see Claus Giering, Flexibilisierungskonzepte für Europa, Arbeitspapier der Forschungsgruppe Europa (München: CAP, 1997); Institut d'Etudes Europeennes (ed.), La differenciation dans l'Union europeenne (Bruxelles: Universite' Libre, 1995). The notion of à la carte Europe was first associated with a famous lecture given by Ralf Dahrendorf, A Third Europe?, Third Jean Monnet Lecture (Florence: European University Institute, 1979), has long remained in the focus of the British intellectual debate - see Helen Wallace with Adam Riley, Europe: The Challenge of Diversity (London, RIIA Chatham House Papers 29, 1985) - and has also become a recurrent trait in the attitude of successive British governments up to the White Paper presented at the start of the latest IGC: see A Partnership of Nations: The British Approach to the EU Intergovernmental Conference 1996, Foreign and Commonwealth Office, Cmnd 3181, London. In a different light see also Pierre Maillet, Dario Velo, L'Europe à Géométrie Variable. Transition vers l'integration (Paris: L'Harmattan, 1994).

7. ⁷ See Alexander C.-G. Stubb, A Categorisation of Differentiated Integration, Journal of Common Market Studies, June 1996, pp.283-29 (I have basically followed his categorisation here); Claus Giering, Vertiefung durch Differenzierung - Flexibilisierungskonzepte in der aktuellen Reformdebatte, Integration, 2/1997, pp.72-8.

8. ⁸ Cf. CDU/CSU-Fraktion des Deutschen Bundestages, Überlegungen zur europäischen Politik. Vorschläge für eine Reform der Europäischen Union, CDU-CSU Dokumentation, Januar 1995. Moreover, the paper explicitly mentioned only five EU States - Germany, France and the Benelux countries, i.e. the founding members of the EC minus Italy - as likely participants to the envisaged Kerneuropa, thus triggering hostile reactions all across Europe and further damaging the cause.

9. ⁹ See Centre for Economic Policy Research (eds.), Flexible Integration: Towards a More Effective and Democratic Europe, Monitoring European Integration, 6, November 1995; Christian Deubner, Deutsche Europapolitik: Von Maastricht nach Kerneuropa? (Baden-Baden: Nomos Verlagsgesellschaft, 1995); Claus-Dieter Ehlermann, Différenciation accrue ou uniformité renforcée?, Revue du Marché Unique Européen, 3/1995, pp.191-218; Club de Florence (ed.), L'Europe: L'impossible status quo (Paris: Stock, 1996); Philippe Manin, Jean-Victor Louis (eds.), Vers une Europe differenciee' ? Possibilite' et limite (Paris: Pedone, 1996); Bertelsmann Foundation (ed.), The New Europe - Strategies for Differentiated Integration (Gütersloh: Bertelsmann Foundation Publishers, 1997), esp. pp.42-49; Francoise de La Serre, Helen Wallace, Les coopérations renforcées: une fausse bonne idée? , Études et Recherches, Notre Europe, Paris, 2/1997.

10. ¹⁰ See fn.8 for the CDU/CSU Paper; John Major, William and Mary Lecture, Leiden, September 7, 1994. Cf. also then French PM Edouard Balladur's interview, Le Figaro, August 30, 1994, and his article, Le Monde, November 30, 1994, in which he referred to a Europe of 'concentric circles' - a common pillar for economic policy, political cooperation and common borders, and solidarités renforcées in other areas such as EMU and defence, plus a third circle for applicants and future members. On the French debate cf.. Anne-Marie Le Gloannec, Europe by Other Means ?, International Affairs , 73, January 1997, pp.83-98. For further references see Deirdre Curtin, The Shaping of a European Constitution and the 1996 IGC: 'Flexibility' as a Key Paradigm, Aussenwirtschaft, 50, 1995, pp.237-252; Frank Vibert, Structured Flexibility in the European Union (London: European Forum, 1996); Alexander C.-G. Stubb, The 1996 Intergovernmental Conference and the Management of Flexible Integration, Journal of European Public Policy, 1/1997, pp.37-55. The prospect of enlarging the EU eastwards, too, played a significant role in the debate: see Francoise de la Serre, L'élargissement aux PECO: quelle différentiation?, Revue du Marché Commun, Novembre 1996; Lee Miles, John Redmond, Enlarging the European Union: The Erosion of Federalism ?, Cooperation and Conflict , 31, September 1996, pp.285-309; Stephan Kux, GASP und Beitrittskandidaten: Blockierung, Flexibilisierung oder vernetzte Sicherheitsgemeinschaft ?, Oesterreichische Zeitschrift fuer Politikwissenschaft , 4/1996, pp.413-430; and Peter van Ham, Central Europe and the EU's Intergovernmental Conference: The Dialectics of Enlargement, Security Dialogue, Winter 1997, pp.71-82; and Gunilla Herolf (ed.), EU Enlargement and Flexibility (Stockholm: The Swedish Institute of International Affairs, 1998).

11. ¹¹ More specifically, Italy proposed QMV should be the rule also for setting general foreign policy guidelines, supplemented by the use of "constructive abstention". If these changes would be accepted by the IGC, Italy saw no need for a flexibility option, which had rather to be considered - consistently with Italy's overall vision - as a second best solution very close to a last resort ('extrema ratio'). As such, however, it could be applied to defence matters, provided participating states included all WEU members. Conversely, the Portuguese proposal argued - as far as the second 'pillar' is concerned - that enhanced cooperation should apply only to the implementation of measures unanimously decided by the Council. For a complete documentation on the 1996-97 IGC cf. the Website http://europa.eu.int/en/agenda/igc-home/index.html .

12. ¹² For early assessments see Kirsty Hughes and Edward Smith, New Labour - New Europe?, International Affairs , January 1998, pp.93-104; Alexander Kazamias, The Quest for Modernisation in Greek Foreign Policy and Its Limitations, Mediterranean Politics, Autumn 1997, pp.71-94. See also Stephen George (ed.), Britain and the European Community: The Politics of Semi-Detachment (Oxford: Clarendon Press, 1992); Heinz-Jürgen Axt (ed.), Greece and the European Union: Stranger among Partners? (Baden-Baden, Nomos Verlagsgesellschaft, 1997); and Lionel Barber, Britain and the New European Agenda, Research and Policy Paper Notre Europe, Paris, 4/1998.

13. ¹³ See Josef Janning, Dynamik in der Zwangsjacke - Flexibilität in der Europäischen Union nach Amsterdam, Integration, 4/1997, pp.285-291; Id., Differenzierung als Integrationsprinzip: Die Flexibilitaet im neuen EU-Vertrag, in Werner Weidenfeld (Hg.), Amsterdam in der Analyse (Guetersloh: Verlag Bertelsmann Stiftung, 1998), pp.203-217. For a detailed reconstruction of the IGC and a first assessment cf. Alexander C.-G. Stubb, The Amsterdam Treaty and Flexible Integration: A Preliminary Assessment, Paper presented at the IPSA meeting, Brussels, 10-12 July 1997; Geoffrey Edwards, Alfred Pijpers (eds.), ThePolitics of European Treaty Reform - The 1996 Intergovernmental Reform and Beyond (London: Pinter, 1997); Geoffrey Edwards, Eric Philippart, Flexibility and the Treaty of Amsterdam: Europe's New Byzantium, in Centre for European Legal Studies, Flexibility and the Treaty of Amsterdam: Europe's New Byzanthium? (Cambridge: CELS Occasional Papers No.3, 1997), pp.1-46. For a more policy-oriented analysis cf. Christian Deubner, Die verstaerkte Zusammenarbeit im System der Europaeischen Union , SWP, AP 3064, Ebenhausen, Maerz 1998.

14. ¹⁴ For recent assessments see - in increasing order of criticism - Elfriede Regelsberger, Mathias Jopp, Und sie bewegt sich doch! Die gemeinsame Aussen- und Sicherheitspolitik nach den Bestimmungen des Amsterdamer Vertrages, Integration , 4/1997, pp.255-263; Philippe de Schoutheete, L'avenir de l'Union Européenne, Politique Étrangère, 3/1997, pp.263-277; Franco Algieri, Die Reform der GASP - Anleitung zu begrenztem gemeinsamen Handeln, in Werner Weidenfeld (Hg.), Amsterdam in der Analyse (Guetersloh: Verlag bertelsmann Stiftung, 1998), pp.89-120; Joerg Monar, The European Union's Foreign Affairs System after the Treaty of Amsterdam: A Strengthened Capacity for External Action? , European Foreign Affairs Review , 2, 1997, pp.413-436; Brian L.Crowe, Some Reflections on the CFSP, European Foreign Affairs Review , 3, 1998, pp.319-324; Nicole Gnesotto, Defense europeenne et partenariat atlantique, in Francoise de la Serre, Christian Lequesne (dir.), Quelle Union pour quelle Europe ? L'apres-traite' d'Amsterdam (Bruxelles: Editions Complexe, 1998), pp.67-95; Philip H. Gordon, Europe's Uncommon Foreign Policy, International Security, Winter 1997/98, pp.74-100.

15. ¹⁵ Roughly speaking, the first phase would coincide with the practical measures to take place after the signing of the Maastricht Treaty revision. During the second phase, the WEU Secretariat would be incorporated into the EU Council Secretariat and the EU would take over the political control of WEU. The third phase would entail the disappearance of WEU as an independent organisation and the establishment of direct relations between EU and NATO.

16. ¹⁶ Actually, so far [early September 1998], the 'click-in' device enshrined in Art.J.4 TEU has been put into effect only once, in November 1996, when the EU called on WEU for the possible organisation of a humanitarian operation in the region of the African Great Lakes (which eventually did not take place). The policing operation in the EU-administered city of Mostar, in former Yugoslavia, was carried out (Summer 1994-Fall 1996) by WEU on the basis of a bilateral memorandum of understanding with the EU.

17. ¹⁷ On their attitudes see Thomas Pedersen, The European Union and the EFTA Countries: Enlargement and Integration (London: Pinter, 1994); Paul Luif, On the Road to Brussels: The Political Dimension of Austria's, Finland's and Sweden's Accession to the European Union (Vienna: Braumüller, 1995), especially pp.237 ff.; Lee Miller (ed.),The European Union and the Nordic Countries (London: Routledge, 1996); and John Redmond (ed.), The 1995 Enlargement of the European Union (Aldershot: Ashgate, 1997). Since 1995, however, such attitudes have increasingly been put into question: in Austria, for instance, the present 'Grand Coalition' government is split over the prospect of future NATO membership. Furthermore, Austria and Finland are pushing for closer integration between EU and WEU, and even Sweden is considering closer operational integration with NATO forces. At any rate - formally speaking, at least - NATO membership is not a prerequisite for WEU full membership.

¹⁸ Cf. Simon Nuttall, The CFSP Provisions of the Amsterdam Treaty: An Exercise in Collusive Ambiguity, CFSP-Forum , 3/1997, pp.1-3. See also the conclusions drawn by Eric Remacle, La PESC et l'UEO face aux directoires, coalitions de volontaires et geometries variables institutionnelles, in Mario Telo', Eric Remacle (eds.), L'Union Europeenne apres Amsterdam. Adaptations institutionelles, enjeux de la differenciation et de l'elargissement , Rapport remis a la Fondation Paul-Henri Spaak, Bruxelles, janvier 1998, pp,82-96.

19. ¹⁹ Actually, the French proposal does not aim at a modification of the Brussels Treaty (1948), whose Art.VIII already envisages non-unanimous decisions, albeit indirectly ('The Council shall decide by unanimous vote questions for which no other voting procedure has been or may be agreed). Moreover, France - backed by other some WEU members - only aims at applying such consensus aménagé to non-Art.V, "Petersberg-type" missions. No formal decision has been taken so far on the matter. The WEU ministerial meeting held in November 1997 in Erfurt, however, emphasised the need for easing consensus-building inside the organisation, if necessary also by exempting some member/associate States from financially contributing to a specific action they would be willing to support politically. Ways are being sought to put all this into practice.

20. ²⁰ For further particulars cf. - respectively B 'La France s'engage a deployer jusqu'a 50 000 hommes pour l'Alliance Atlantique', Le Monde , 2 juillet, 1998; 'Une armee francaise otanisee', Le Monde , 3 juillet, 1998; Jonathan Eyal, Defense on the Cheap, Financial Times , July 9, 1998; 'New Model Armies, Time , July 27, 1998; Lawrence Freedman, New World, New Roles, ibid. ; and 'Pan-Europe Defense Group A Step Closer', Financial Times , July 11/12, 1998; 'GEC and Alenia in Joint Venture', ibid. ; 'Sweden Hints at Sale of Celsius', Financial Times , July 14, 1998; 'France in Policy Change with Aerospatiale Merger', Financial Times , July 23, 1998; 'Nordic States Move Towards Joint Defense Purchases, Financial Times , August 8, 1998; 'Boost for European Weapon Agency', Financial Times , September 10, 1998.

21. ²¹ According to Duchene, "the EC's interest as a civilian group of countries long on economic power and relatively short on armed force is as far as possible to domesticate relations between States, including those of its own members and those with States outside its frontiers. This means trying to bring to international problems the sense of common responsibility and structures of contractual politics which have been in the past associated almost exclusively with 'home' and not foreign, that is alien , affairs" [quot. from Francois Duchene, The European Community and the Uncertainties of Interdependence, in Max Kohnstamm, Wolfgang Hager (eds.), A Nation Writ Large ? Foreign-policy Problems before the European Community (London: Macmillan, 1973), pp.19-20]. More recently, Hanns Maull has defined "civilian power" as involving "a) the acceptance of the necessity of cooperation with others in the pursuit of international objectives; b) the concentration of non-military, primarily economic means to secure national goals, with military power left as a residual instrument serving essentially to safeguard other means of international interaction; and c) willingness to develop supranational structures to address critical issues of international management" [quot. from Hanns W.Maull, Germany and Japan: The New Civilian Powers, Foreign Affairs , 69, 5/1990, pp.92-93].

22. ²² In a seminal article dating back to the mid-1960s, Mancur Olson and Richard Zeckhauser developed an economic theory of alliances by characterizing deterrence as a 'public good': Mancur Olson, Richard Zeckhauser, An Economic Theory of Alliances, Review of Economics and Statistics , 48, 1966, pp.266-279. The purely 'public good' nature of defense in general has been increasingly put into question ever since, in particular by Todd Sandler - see Todd Sandler, Impurity of Defense: An Application to the Economics of Alliances, Kyklos, 30, 3/1977, pp.443-460; Todd Sandler, Jon Cauley, The Design of Supranational Structures, International Studies Quarterly , 21, 2/1977, pp.251-276; and Todd Sandler, John F.Forbes, Burden Sharing, Strategy, and the Design of NATO, Economic Inquiry , 18, 1980, pp.425-444 - who argued that defence has become an 'impure' public good and that the costs and benefits of forming tight cooperative structures depend not only on economies of scale but also on transaction costs. See also John R.Oneal, The Theory of Collective Action and Burden-Sharing in NATO, International Organization , 44, 1990, pp.379-402; and Mark A.Boyer, International Cooperation and Public Goods: Opportunities for the Western Alliance (Baltimore: Johns Hopkins University Press, 1993). For an excellent review of the literature and some interesting suggestions for further research - encompassing also such variables as the nature of threats, the specificity of the assets involved and the degree of heterogeneity of participating States - see Katja Weber, Hierarchy Amidst Anarchy: A Transaction Costs Approach to International Security Cooperation, International Studies Quarterly , 41, 1997, pp.321-340. Meanwhile, more specific analyses of 'clubs' and 'cores' have been developed in the field of public choice - see e.g. Dennis C.Mueller, Perspectives on Public Choice: A Handbook (Cambridge: Cambridge University Press, 1997); Richard Cornes, Todd Sandler, The Theory of Externalities, Public Goods and Club Goods (Cambridge: Cambridge University Press, 1996) - and game theory: see e.g. Robert Axelrod, The Complexity of Cooperation: Agent-Based Models of Competition and Collaboration (Princeton: Princeton University Press, 1997).

23. ²³ For one thing, since 1993 the duration of the WEU rotational presidency has been reduced from one year to six months, in accordance with the provisions laid down in the Maastricht Treaty for the EU presidency. Secondly, from January 1, 1999 - starting with Germany - the rotational presidencies of the EU and WEU will be held jointly whenever a WEU full member takes over the EU Council. The decision was taken by WEU after the signature of the Amsterdam Treaty. No trojka - either Maastricht- or Amsterdam-type - nor any similar mechanism, however, has (yet) been established for WEU.

24. ²⁴ On these aspects see Stephan de Spiegeleire, From Mutually Assured Debilitation to Flexible Response: A New Menu of Options for European Crisis Management, in Guido Lenzi (ed.), WEU at 50 (Paris, WEU Institute for Security Studies, 1998), pp.15-34.

25. ²⁵ The concept of "framework nation" (nation-cadre) was originally a French proposal. Under the more usual name of "lead nation", however, it already enjoyed wide applicability in the United Nations and in other international organisations that face similar difficulties with collective action problems. It was formally adopted in the Paris Declaration of the WEU Council of Ministers on May, 13, 1997. It applies to the organisation of autonomous WEU operations, of which it is a special case (it allows for a faster setting up of an operation by using a national headquarters, while at the same time emphasising its multinational nature by ensuring broad coalition representation on the lead-nation headquarters staff) - and is Adesigned to enable a European Headquarters to be established, using existing national or multinational assets, within timeframes compatible with the operational requirements, especially in situations of extreme urgency. It explicitly seeks to envisage "flexible modes of action that are adaptable to a range of crisis situations" (italics added).

26. ²⁶ For a first assessment see Stefano Silvestri, The Albanian Test Case, The International Spectator, July-December 1997, pp.87-98; Franck Debie, La Grèce, l'Italie et l'Europe face au problème albanais. Gestion de crise et représentations géopolitiques, Relations Internationales & Stratégiques, 28, Hiver 1997, pp.96-108; Georgios Kostakos, Dimitri Bourantonis, Innovations in Peacekeeping: The Case of Albania, Security Dialogue, 29, March 1998, pp.49-58; Ettore Greco, New Trends in Peacekeeping: The Experience of Operation Alba, Security Dialogue , 29, June 1998, pp.201-212; and Edward Foster, Ad Hoc in Albania: Did Europe Fail ?, Security Dialogue , 29, June 1998, pp.213-217. For a tentative connection between conflict prevention and crisis management, as well as between second and third pillar policies, see also Ferruccio Pastore, Conflicts and Migrations: A Case Study on Albania , CeSPI Occasional Papers, March 1998. It is worth mentioning, too, that in the wake of 'Alba' the WEU Council formally pledged to give primacy from now on - on similar occasions - to multilateral operations undertaken within the W/EU institutional framework over looser ad hoc coalitions.

27. ²⁷ For a thorough analysis see Joerg Monar , Roger Morgan (eds.), The Third Pillar of the European Union (Bruxelles: College de l'Europe, European Interuniversity Press, 1995); Michael Anderson et al., Policing the European Union (Oxford, Clarendon Press, 1996); and especially Didier Bigo, L'Europe de la securite' interieure: penser autrement la securite', in Anne-Marie Le Gloannec, Entre Union et Nations. L'Etat en Europe (Paris: Presses de Sciences Po, 1998), pp.55-90, and Id., Les non-dits de la securite' interieure, in Francoise de la Serre, Christian Lequesne (dir.), Quelle Union pour quelle Europe ? L'apres-traite' d'Amsterdam (Bruxelles: Editions Complexes, 1998), pp.43-65.

28. ²⁸ This is an obvious simplification justified only by the lack of time. Several interesting techniques have been put aside like modelling, spectral analysis, morphological analysis, …

29. "the analyst may decide not to attempt to optimize but rather to 'satisfice' -that is, select a solution that may not be the best, but one that is good enough and that can be agreed upon. Perhaps the majority of policy solutions, especially quick solutions, are arrived at through satisficing. In fact, optimizing when there are conflicting or multiple goals often is extremely difficult. In some cases there may be no way to combine or add up the various impacts.". Patton, Carl V., Sawicki, David S. (1993), Basic Methods of Policy Analysis and Planning, Englewood Cliffs (N-J): Prentice Hall, p.337.

30. ³⁰ previsions built on an opinion survey of a population of experts; answers collected individually through questionnaire and global results forwarded to each expert who are then asked to answer again, and so on in order to arrive at a stabilization of the answers; majority opinion

31. ³¹ The European defence industry sector has often been seen as the locus of functional, political, cultivated and even 'reactive spillover' since the mid-1980s (Rem.: 'reactive spillover' refers to the fact that "spillover pressures were sparked by change in international politics according to …' changes in the international political and security environments were critical components" Guay T. R. (1998) At arm's length - The European Union and Europe's defence industry London: MacMillan p.177). There is a growing realization that, individually, EU member states and national defence champions are too small to survive and that European cooperation has to be quickly organized. In the declaration on WEU attached to the Maastricht Treaty, the examination of 'enhanced cooperation in the field of armaments with the aim of creating a European armaments agency' was already listed among various ways to strengthen WEU's operational role. New TEU Article 17 goes a bit further by introducing the idea in the core of the Treaty, stating states that 'The progressive framing of a common defence policy will be supported, as Member States consider appropriate, by cooperation between them in the field of armaments.'

32. ³² Hybrid cooperation with the creation in 1992 of the Western European Armaments Group (WEAG) part of the WEU, which has the responsibility to negotiate the principles and operational modes of a future European armaments agency EAA; slow development of a (lax) common regime of cooperation with the adoption of principles in 1994. Intra-EU cooperation with the creation in July 1995 of a working group for armament policy (POLARM) studying the principles of a 'future European technological and industrial defence policy'. Extra-EU ad hoc formula with the creation by France and Germany in January 1996 of a specific (bilateral) structure to manage their bilateral armaments programmes; the structure was then enlarged to the UK and Italy, to form the Organization of Cooperation in armaments (OCAR) in November 1996; illustration of the anxiety of the smaller member states not to be kept out of an otherwise disruptive initiative in terms of European integration, the Netherlands, Belgium and Spain have asked to join…(see de Vestel P. (1998), 'The future of armament cooperation in NATO and the WEU', in Eliassen Kjell A. (ed.), Foreign and Security Policy in the European Union, London: Sage Publications, pp.197-215).

33. ³³ On the costs and benefits of the two main options in design strategy, i.e. an inclusive strategy and a sequential strategy, see the discussion by Downs et alii on supply side of multilateralism in International Organization. "Instead of forming an 'inclusive' agreement - that is, one that covers nearly all of the states that its designers eventually hope to include - many multilateral organizations start out with substantially smaller memberships and generally expand over time. We argue that, among the many possible explanations for the choice of this design strategy, there is a rational choice argument that contains behavioral implications not found in other theories. We show how the strategy of admitting potential members sequentially over time based on their preferences for cooperation is able to generate endogenously a series of structure-induced equilibria. The resulting path-dependent process produces a multilateral organization that will often be deeper at every stage of its development than would be obtained by an inclusive strategy; and it mitigates, even if it does not fully eliminate, the breadth-depth trade-off so prominent in the existing literature. As a result, large multilaterals that start out small will tend to become considerably "deeper' in a cooperative sense than those that start out with many members. This outcome holds whether or no sequential growth has been pursued for strategic reasons. As long as those states favoring deeper cooperation tend to be admitted before more conservative states (that is, those desiring less coopération), the efficiency advantages of this design strategy should hold." Downs George W., Rocke David M., and Barsoom Peter N. (1998), "Managing the Evolution of Multilateralism", International Organization, Vol.52, No. 2, Spring, pp.397-419.

34. ³⁴ Quade identifies "the five most important elements in the policy analysis process as problem formulation, searching for alternatives, forecasting the future environment, modeling the impacts of alternatives, and evaluating (comparing and ranking) the alternatives. He points out that policy analysis is an iterative process in which the problem is reformulated as objectives are clarified, alternatives are designed and evaluated, and better models are developed. ". Quade Edward S. (1982), Analysis of Public Decisions, 2^nd ed. New York: Elsevier Scientific, pp, 47-62.

35. ³⁵ Helen Wallace argues against the false 'magic of numbers' and the necessary adjustments imposed by future enlargements. She contends that, when the 'old' members evoke the need for an institutional antidote, the fundamental question is less about 'numbers' than about their relative position. According to her, the level of flexibility which would eventually be needed might be far lower than what is currently thought (Wallace 1999).

36. ³⁶ In June 1998, at the French national Assembly, the French Minister for European Affairs, Pierre Moscovici, stated clearly that the next enlargement 'raises, no doubt for the first time, the question of the limits of European construction, not only its geographical but also its "functional" limits ... This does not imply any infringement of the principle of solidarity, nor any manoeuvre by the older MS against the new ones: it is simply a virtual biological necessity in a Europe of 20 or 30' (Agence Europe, 6 June 1998).

37. The following remarks are based also on a series of specific auditions held by an international panel in the Netherlands, Germany, Italy, Spain, Poland and Denmark.

38. The terms "institution" and "organization" will be used interchangeably, i.e., "organization" should not be understood in the strict legal sense as an IGO, excluding borderline cases like the OSCE.

39. It is worth noting that the word „beeinträchtigen" in the final German text of the Amsterdam Treaty is restricted to the negative consequences of someone else's behaviour. It replaced the verb "berühren", which is the German equivalent of "affect", in earlier drafts.

40. Hence the term "destructive" (as compared with "normal") abstention might be more appropriate. Simon Nuttall, The CFSP Provisions of the Amsterdam Treaty: An Exercise in Collusive Ambiguity , CFSP Forum 3 (1997), p. 1-3 (p. 2); for further details, see Hanspeter Neuhold, The Provisions of the Amsterdam Treaty on the CFSP: Cosmetic Operation or Genuine Progress? in: Gerhard Hafner et alii (eds.), Liber Amicorum: Professor Ignaz Seidl-Hohenveldern in honour of his 80^th birthday (The Hague/London/Boston 1998), p. 495-510, and the literature quoted there. It should also be noted that "constructive" abstention is not applicable to decisions having military or defence implications (Art. 23, para. 2, subpara. 4 TEU).

41. Finland's and Sweden's non-membership in military alliances should not be confused with the policies of the Non-Aligned Movement mainly composed of countries from the "Third World".

42. Other examples of internal flexibility within the EU are the Eurocorps as well as EUROFOR and EUROMARFOR.

43. However, the WEU remains a separate, independent international organization. But since all members of the WEU participate in the adoption of guidelines by the European Council, they are estopped from deviating from these guidelines later on within the WEU. Christoph Thun-Hohenstein, Der Vertrag von Amsterdam. Die neue Verfassung der EU (Vienna 1997), p. 68.

45. It makes a difference whether a state would like to join a flexible initiative but is unable to do so or whether a state which has the ability to take part refuses participation because it does not accept the costs and risks involved.