A common european export policy for defence and dual-use items

*This paper was first published as Working Paper n° 9 by the Study Group on Enhancing Multilateral Export Controls for US National Security (The Henry L. Stimson Center/CSIS, April 2001).

In Europe, arms and dual-use exports raise complex questions. First, they fall between two policy spheres that are organised in a distinctly contrasting manner. On the one hand, they are an intrinsic part of commercial policy that lies within the exclusive competence of the European Community (EC). On the other hand, they come under the aegis of security and defence policy, a jealously guarded area of responsibility of the EU member states. Second, there is no common European export policy. Arms exports, in particular, remain a sensitive issue on which Europeans have difficulties in reaching a consensus. Strong divisions exist mainly (but not exclusively) between arms producing and non-producing countries.

Consequently, the institutional setting is complicated. Whilst the dual-use regime forms part of pillar one of the European Union (EC), the Code of Conduct on arms exports belongs to pillar two (CFSP). Moreover, the six major arms producing European countries have established their own rules on transfers and exports for cooperative programmes that remain completely outside of the EU framework (the Letter of Intent, or LoI).

The European Community export control regime for dual-use goods is based on mutual recognition of national export decisions. Moreover, the Council Regulation in question is part of Community law. As a consequence, multilateral non-proliferation agreements are legally binding on all EU member states. The EU Code of Conduct for Arms Exports, in contrast, provides only broad moral and political standards. It is not legally binding and its consultation mechanism is mainly restricted to undercutting. Finally, the Framework Agreement between the major arms producing countries establishes an export regime for cooperative armaments projects that is, in effect, a de facto licence-free zone among the six participants and a decision-making mechanism for exports of jointly developed weapon systems to third countries. In contrast to the dual-use regime, export provisions of the Framework Agreement operate only on an ad hoc basis and are applied differently to each and every cooperative programme.

The absence of a common export policy undermines all three arrangements. Indeed, the dual-use regime is built upon the mutual acceptance of divergent policies, whereas the Code of Conduct merely tries to limit the worst excesses of political divergence on arms exports, and the Framework Agreement offers only a mechanism for ad hoc agreements.

Equally, all three arrangements are ongoing processes that might one day form the basis for convergence towards a common policy. With regards to dual-use items, there is a clear trend towards liberalisation that will probably become even stronger following the recent strengthening of the Commission and the introduction of qualified majority voting in the Council. By contrast, the EU Code of Conduct suffers from persistent divergences. Although it has established a consultation process that might, in time, help bridge these gaps, progress toward a truly common export policy will be slow and dependent on the development of a Common European and Security Policy (CESDP) and a Common Foreign and Security Policy (CFSP).

Decision-making under the LoI Framework Agreement promises to be both sensitive and substantive. Focusing on specific export decisions and covering the most important programs, the LoI mechanism goes to the core of European policy divergences, and could become a powerful and effective driving force towards European convergence. This does not mean that a rapprochement will be easy to achieve. Indeed, conflicts between the current German government and its French and British counterparts seem preordained. Nevertheless, the ongoing cross-border restructuring of European defence industries is putting all LoI governments under pressure to harmonize their export policies. The most ‘Europeanized’ sectors, namely aerospace and defence electronics, are actively lobbying for a broad application of the principles of the Schmidt-Debré accord within the context of the LoI process, leaving the final decision to the country that holds the contract. In accordance with the principles of the dual-use regime, such an agreement would allow the free circulation of defence items among the parties concerned. From an economic point of view, this would indeed be the most rational and logical solution. In the foreseeable future, however, such an agreement could, at best, only be reached outside the EU framework among the LoI partners.

Control of arms and dual-use exports is a complex issue in Europe because it lies between two very different policy worlds. On the one hand, it is related to commercial policy, which lies within the exclusive competence of the European Community (EC). On the other, it is linked to security and defence policy, a sphere that remains the national responsibility of the individual EU member states. This raises awkward questions of competence.

For dual-use items, which are primarily intended for civil applications, the process of Europeanization is the most advanced. In order to achieve a truly integrated internal market and a more complete EC commercial policy, the member states established a common system for export controls, where dual-use items may move freely within the Community because all member states recognize each other’s authorization for exports to third countries. There is, however, no common export policy, and in certain areas, national prerogatives and restrictions persist.

A similar regime does not (yet) exist for defence goods. Article 296 of the Treaty of the European Communities excludes military goods from the common market.[1] Hence, each European country has its own cumbersome export legislation and its own export policy. There is neither free movement of defence goods among EU member states nor a common control regime for exports to third countries.

However, even in the pure military arena, several interrelated developments have challenged this national predominance and the lack of harmonization between member states:

The EU has started to develop its own Common European Security and Defence Policy (CESDP) as part of its Common Foreign and Security Policy (CFSP). Armaments issues are not (yet) covered, but are, of course, related to both policies. Hence, there is a political and institutional logic to bring at least some of them into an EU framework.

With the growing importance of commercial technologies for military equipment, the distinction between defence and dual use goods becomes increasingly blurred, challenging the juxtaposition of different regulatory frameworks for each area.

Due to developments both on the supply side (cross-border consolidation of industry) and the demand side (limited budgets, growing need for interoperability), more and more weapon systems in Europe will be developed through international cooperation. This raises thorny questions concerning the transfer of components and subsystems among European nations and the possible exportation of jointly produced systems to third countries.

Starting with aerospace and electronics, defence (-related) industries in Europe are becoming increasingly transnational in nature. Without a homogeneous defence economic space, including common regulations for transfers and exports, transnational defence companies (TDCs) are extremely difficult to operate. The situation for industry is all the more complex since most high tech companies have both civil and military activities and have to operate within two different regulatory frameworks. Hence, there is strong pressure from industry on governments for more harmonization, in particular since it was governments that urged industry to restructure across national borders.

These factors have created a political dynamic that has led to several initiatives:

Although operating in different frameworks (EU first pillar, EU second pillar, outside the EU), all three initiatives are in a certain way complementary. They demonstrate that Europe is in a transition phase where even in the fields of security and defence sovereignty is progressively delegated from the national to the transnational level. It is not surprising that this development starts with commercial aspects of armaments.

CHAPTER ONE: THE COMMON CONTROL REGIME FOR DUAL-USE EXPORTS

The question of export controls on dual-use goods was first raised in the context of the completion of the Internal Market (1992). Considering that dual-use goods should move as freely between member states as they do within each of them, and that control on intra-EC trade could only be eliminated if all member states established effective controls based on common standards for exports to non-EC countries, the Commission submitted a proposal for a Council Regulation on 31 August 1992. On the basis of this proposal, the EU Council adopted on 19 December 1994 a system of export controls on dual-use goods consisting of Council Decision 94/942/CFSP, on the one hand, and Council Regulation (EC) 3381/94, on the other. The regime entered into force on 1 January 1995 and became applicable six months later, on 1 July 1995.[2]

I.1 The 1995 control regime

The combination of Council Decision and Council Regulation was a cross-pillar approach aiming at coping with the responsibility dilemma connected to the specificity of dual-use goods. As part of commercial policy, restrictions on exports of dual-use goods fall within the competence of the Community by virtue of Article 133 (formerly 113) EC.[3] Therefore, control procedures and mechanisms were outlined in Regulation (EC) n° 3381/94, which became part of Community law. The Council Decision, in contrast, was adopted under Article J.3 of the Maastricht Treaty (now Article 14 of the TEU) concerning joint action in matters covered by the CFSP.[4] Both texts were closely entwined by numerous cross-references and formed an ‘integrated system’. The Community Regulation outlined how the regime would work, whereas all lists of destinations and of controlled items were annexed to the Council Decision under CFSP rules. It was the Council’s (and therefore the member states’) sole responsibility to establish, monitor and update the lists. Since the latter are obviously the heart of the control regime, the Regulation alone would have had no substance and made no sense. The member states’ prerogatives were based on the assumption that all decisions concerning the lists were strategic and/or political in nature and, therefore, outside the Commission’s competence.

A common list of dual-use goods requiring a license if exported from the Community (Annex I of the Council Decision). This list was - and still is - a compilation of the control lists defined by other non-proliferation regimes (Wassenaar, MTCR, Nuclear Suppliers’ Group).

Mutual recognition of export licenses granted by the competent authorities of member states (Article 6.3 of the Regulation).

Free movement of dual-use goods inside the Community, with the exception of certain highly sensitive goods listed in Annex IV and V of the Council Decision. Annex IV included those items for which export authorizations were required by all member states; Annex V included lists of dual-use goods for which certain member states maintained national controls on intra-Community transfers. (Hence, the lists compiled in Annex V varied from country to country.)

A catch-all clause which makes any good subject to licensing requirement if the exporter is informed by his authorities or ‘knows’ that a good is intended to be used in relation to a program of weapons of mass destruction (Article 4 of the Regulation).

Some harmonization of export licenses. The Regulation evokes three types of licenses: individual export licenses, granted for the most sensitive exports; global licenses granted to a specific company for exports of certain goods to certain destinations; and general licenses offering simplified procedures for export of controlled goods to certain destinations (Article 6 of the Regulation).

In the absence of a common list of prohibited destinations, it is for competent national authorities to decide on the export of dual-use goods (Article 7 of the Regulation). They shall base their decisions on common guidelines, outlined in Annex III of the Council Decision.[5]

For exports to certain allied and friendly countries (included in Annex II of the Council Decision) the general authorization applied upon mere request of the exporters.[6] However, this provision was not exclusive; for other destinations, the competent authorities could grant the same facility of simplified formalities to exporters who requested it. On the other hand, member states had the possibility to deny general authorizations for items listed in Annex IV and V.

Article 22 provided that Article 296 EC and the Euratom Treaty[7] would remain unaffected. This means that purely military goods included in the list adopted by the Council under Article 296 EC and nuclear materials coming under the Euratom Treaty (but not nuclear products and technology covered by the EC Treaty) are not covered by the dual-use Regulation.

I.2 The review process

In July 2000, Council Decision 94/942/CFSP and Regulation (EC) 3381/94 were replaced by Council Decision 2000/402/CFSP and Regulation (EC) 1334/2000, reforming the Community regime for the control of dual-use exports.[8] The reasons for this change were both legal and practical.

First, the integrated system was a legally doubtful construction. According to the European Court of Justice (ECJ), the legal basis of Regulation 3381/94 was Article 133 EC alone, since it was restricted to matters that fall entirely within the sphere of Community exports and customs policy. In this regard, the fact that export controls were established for reasons of foreign and security policy did not change anything. This judgment, in turn, was incompatible with the fact that the Regulation could not work without Council Decision 94/942/CFSP. In fact, it was the Council, i.e. the member states, that decided on the specific lists of items and countries to which the Regulation applied. This construction did not only devaluate the position of the Commission, which normally has the monopoly for initiating Community law, but was also in contradiction to the principles of the EC treaty. The Council Decision, through its integration with the Regulation, came into the jurisdiction of the ECJ (which normally has no say on CFSP matters), thereby raising more legal problems. The only way to overcome these contradictions was to base the whole export control regime exclusively on Article 133.[9]

The second set of reasons to change the system was practical. Article 18 of the Regulation stipulated that the Commission should present after two years a report to the European Parliament and the Council on the application of the Regulation. This report, published in 1998, identified a number of deficiencies and problems, concerning mainly the licensing system and the catch-all clause.

According to the report, the practical application of the principle of mutual recognition suffered from the absence of standardized license forms and of harmonized licensing procedures. The system was far too complex to be routinely managed by customs officials at the border, and too cumbersome to be useful for industry in practice. For general licenses, for example, the scope of products covered and the destinations allowed for exports differed from country to country.

Moreover, general licenses were, in some cases, defined by legislation only and not materialized by a license document. Global licenses, in turn, represented a new concept for some member states, which has, therefore, not always been understood and accepted by custom officials. The result was considerable delays in cross-border transfers, deterring industry from using the EU regime to export from a given member state with a license issued by another member state. The discrepancy between national licenses was particularly difficult to manage for companies established in several member states. For individual licenses, the situation improved partially due to a Standard Model developed by the Commission and informally used by most member states. For global and general licenses, however, numerous types of national licenses continued to exist. The Commission therefore recommended a) to make the standard model for individual licenses mandatory; b) to harmonize the forms of national licenses; c) to introduce a harmonized Community license for exports to allied and friendly countries listed in Annex II; and d) to require all member states to offer the possibility of global licenses.

Implementation of the catch-all clause - which subjects non-listed dual-use goods to a license requirement if there is a proliferation risk associated with their export - created problems as well. Since the clause was an innovation for most member states, practical application differed widely from country to country. According to the Commission, the main problem was the different degree to which governments inform their exporters about sensitive end-users. This diversity raised questions of distortion of competition and put the effective enforcement of the catch-all clause in doubt. The Commission therefore recommended that member states ‘improve significantly their information-sharing on sensitive end-users with a view to ensuring that a similar degree of guidance is given to exporters throughout the Community.’[10] Finally, the Commission criticized certain practices of the administrative cooperation and stipulated greater liberalization of intra-Community transfers.

I.3 The 2000 control regime

The regime established in June 2000 is an updated version of its 1995 predecessor, and thus not completely new. Modifications are based both on the ECJ’s case law and the Commission’s review.

First, the regime is now no longer an ‘integrated system,’ but is based solely on an EC act. The Regulation (EC) 1334/2000 includes all Annexes (which are organized differently compared to those in the former Regulation) and is effective without any cross-references to the Council Decision 2000/402/CFSP.

Second, Regulation 1334/2000 is more comprehensive and detailed in its definitions than its predecessor. Whereas the former Regulation covered only dual-use ‘goods,’ the new Regulation is on ‘dual-use items, including software and technology [...] and all goods which can be used for both non-explosive uses and assisting in any way in the manufacture of nuclear weapons or other nuclear devices.’ The narrow definition in Regulation 3381/94 was due to the fact that the EC’s commercial policy only covers goods but not services. It soon became clear, however, that technology transfer could not be limited to physical transfer and that the revolution in electronic information transmission had to be taken into account. Accordingly, the new Regulation defines export as ‘transmission of software or technology by electronic media, fax or telephone to a destination outside the Community’ (Article 2.2).

Third, the catch-all-clause has been extended and specified. The 1995 version only obliged an exporter to ask for an export authorization if he was informed by his authorities that the goods in question ‘are or may be intended, in their entirety or in part, for use in connection with’ weapons of mass destruction and ‘missiles capable of delivering such weapons, as covered by the corresponding non-proliferation arrangements’ (Article 4). The new Regulation takes on this provision (dropping the reference to the non-proliferation arrangements) and adds, in Article 4.2, that an authorization is also required for non-listed dual-use items, if (a) the purchasing country or country of destination is subject to an arms embargo imposed by the EU, the OSCE or the UN, and (b) the exporter has been informed that the items in question are or may be intended, in their entirety or in part, for a military end-use. Moreover, Article 4 gives a precise definition of military end-use.[11]

Fourth, the new Regulation harmonizes the licensing system along the lines of the Commission’s proposals:

Article 6.1 of the new Regulation establishes a Community general authorization for most dual-use exports to allied and friendly countries (listed in Part 3 of Annex II).[12]

This means that the European Community becomes a licensing authority in its own right. The Community General License covers all items listed in Annex I, except certain very sensitive items (all items included in Annex IV and certain items covered by the MTCR list, as specified in Part 2 of Annex II). The Community General License may, however, not be used if the conditions of the catch-all clause are met (see above).

For all other exports for which an authorization is required, a license - be it individual, global or general - shall be granted by the competent authorities of the member states (Article 6.2). Several provisions aim at harmonizing these national authorizations:

- Article 6.5 stipulates that member states maintain or introduce into their respective national legislation the possibility of granting a global authorization to a specific exporter for certain items which may be valid for exports to one or more specified countries.

- Article 10 claims that all individual and global export authorizations shall be issued on forms consistent with a standard model (set out in Annex III a).

- General export authorizations granted by national authorities shall be issued in accordance with common guidelines set out in Annex III b. Like their Community counterparts, national general licenses shall not be granted, if the criteria for the catch-all clause are fulfilled (Article 10.2).

Fifth, intra-Community transfers and exports to Annex II countries have been considerably liberalized. Annex V and its various national lists have disappeared completely, and Annex IV has been reduced notably (although certain items of the former have been included in the latter). All in all, the number of items that need authorization for intra-Community transfer has been reduced by two thirds. Moreover, the new Regulation will offer greatly simplified application procedures for exports to allied and friendly countries. In fact, the Community General License implies an almost complete liberalization of exports of about 95 percent of all dual-use items to destinations listed in Annex II. For these items and these destinations, member states have to grant a Community General License to exporters who request it and can no longer impose an individual license.

Sixth, consultation mechanisms and transparency have been improved. The common guidelines on which national authorities shall base their export decisions have been moved from the Annexes to the text of the Regulation (Article 8). Appearing in a more ‘prominent’ place, they gain in political weight but remain vague. The most important innovation in this context is probably the inclusion of a consultation mechanism on undercutting. According to Article 9.3, any member state, before granting ‘an export authorization which has been denied by another member state for an essentially identical transaction within the previous three years, [...] will first consult the member state or states which issued the denial(s). If following consultations, the member state nevertheless decides to grant an authorization, it shall inform the other member state and the Commission, providing all relevant information to explain the decision.’

I.4 Assessment

The 2000 control regime represents, most of all, a strengthening of the Commission vis-à-vis the Council and the member states. The fact that the regime is now based solely on a Community act means that both principles and lists come under the Commission’s prerogative, thus changing completely the regime’s underlying philosophy. The Commission now has the exclusive right of initiative, and all Council decisions will be taken by qualified majority (instead of unanimity). The combination of both might, in the medium term, lead to a situation where the member states lose their exclusive authority to grant or to refuse an export license. Concerning the control lists, member states will continue to have a say as long, as the common list is a compilation of lists drawn up in international fora where they - and not the Commission - are represented. On the other hand, this is more a de facto than a de jure right, and it might only be a question of time for the Commission to ask for representation in the corresponding international bodies. The problem is that the Commission so far does not have the technical competence to assess security aspects of dual-use technologies. This argument, however, can work both ways: either member states will be able to use it against the Commission to maintain their prerogatives or the Commission will use it to its own advantage, claiming the creation of such competence under its own roof (which might open the door for the Commission into the military field as well).

The updated regime provides for a far-reaching liberalization of intra-Community trade by limiting the member states’ ability to impose restrictions to a strict minimum. The same is true for exports to friendly and allied nations. Moreover, the combination of qualified majority voting and the Commission’s exclusive right of initiative makes future liberalization steps probable. In fact, at the time the new export control regime was negotiated, most member states wanted to follow the Commission in its claim to push liberalization further, but the need for unanimity stopped these initiatives. France was the only country, for example, that insisted on treating exports of Annex IV items on a case-by-case basis, whereas others (Germany in particular) suggested that general licenses could be used for all dual-use items exported for civil applications. Now that decisions are taken by qualified majority voting there is a real chance that Annex IV will be further reduced or even abolished. Indeed, some experts already fear that the new single pillar structure of the regime might lead to a situation where commercial considerations will prevail over security considerations.

From the point of view of exporters, liberalization is, of course, an advantage of the new Regulation regime. But there are more benefits to industry. First, the updated system strengthens the legal position of companies in the event of a dispute, since all parts of the regime now come under community law and therefore within the jurisdiction of the ECJ. Second, the licensing system has been considerably harmonized, which should make the regime more workable and, therefore, more attractive to companies.

The catch-all clause, in contrast, continues to be a problem for industry. The new version is more specific, but it still places a huge responsibility on companies. It offers a more detailed description of dual-use items, but it remains difficult to handle. The growing importance of commercial items for military systems will make it increasingly difficult to respect the clause. Moreover, the Regulation does not provide for an improvement in the information exchange between governments and industry. On the other hand, it will be difficult for governments to prove that an exporter was aware that an item might have been intended for a military end-use. The shift from dual-use goods to dual-use items will raise similar problems: it is important to cover technology transfers by new communication means, but in practice, it will be very difficult, if not impossible, to control transfers (via e-mail, etc.). In this regard, the Regulation itself is only a declaration of intent and principles. It remains to be seen how the latter will be implemented by national authorities.

The new consultation mechanism on undercutting will certainly improve cooperation between EU countries. Up to now, a member state could easily grant an export license for an item for which the authorities of another member state had refused authorization. Under the new Regulation, such undercutting can provoke considerable peer pressure. Member states now have to a) inform each other on denials of export licenses; b) consult with each other on their intention to undercut; and c) explain their decision to do so. This provision is a (highly) upgraded version of the consultation mechanism of the Code of Conduct for arms transfers. The fact that it has been possible to include multilateral notification in the dual-use regime but not in the operative provisions of the Code of Conduct shows the different levels of harmonization and integration in the two areas (see Part 2 of the Regulation).

In the medium turn, the new consultation mechanism might even help overcome the main problem of the control regime - that is, the absence of a common export policy. Today, the regime constitutes nothing more than a common framework for different national policies. Member states recognize each other’s export licenses, but they do not necessarily agree with each other’s export policies. The common guidelines remain so vague that there is a real risk of inconsistent interpretation and application of the provisions by national authorities. Increased cooperation among national governments cannot compensate fully for the absence of a common policy, since any intergovernmental process has inevitably its limitations.[13]

A potential source of divergence is, for example, that the catch-all clause refers in its definition of military end-use to national lists. Since these lists differ considerably, misinterpretations are inevitable. The recent creation of a common military list in the context of the Code of Conduct (see below) may improve the situation, but the coexistence of several lists (dual-use/ military, national/ common) creates a gray zone that is all the more difficult to handle since the control regimes, to which the various lists are connected, are organized within different frameworks (dual-use = European Community, military exports = national and CFSP, see below). The more the distinction between military and civil technology gets blurred, the more difficult it will become to implement export controls for dual-use items in a coherent way without harmonizing arms export controls as well.

CHAPTER TWO: THE CODE OF CONDUCT

The European Union Code of Conduct on Arms Exports was adopted on 8 June 1998 by the General Affairs Council as a Council Declaration in the framework of the CFSP. This means that the Code of Conduct is part of the second pillar (intergovernmental) and does not involve the European Commission (normally in charge of trade issues). The initial proposal for the Code was tabled by France and Great Britain in late January 1998 and then discussed in several meetings of the Council’s Working Group on Conventional Arms Exports (COARM) as well as in the Political Committee.

The Code of Conduct sets common minimum standards for the management and control of conventional arms exports by member states to third countries. Moreover, it establishes an information exchange and consultation mechanism, the first ever applied by any group of states in this field.

The overall objective of the Code is to achieve greater transparency in arms transactions and to lead to a growing convergence of national export policies. It is composed of two parts: (1) guidelines that set out a number of circumstances in which licenses should be denied; and (2) operative provisions that contain a mechanism for consultation on undercutting and an annual review process.

II.1 Export criteria

The Code’s eight export criteria are based on those defined by the European Councils in Luxembourg (29 June 1991) and Lisbon (26-27 June 1992). These include:

1. Respect for the international commitments of EU members, in particular the sanctions decreed by the UN, the EC, and non-proliferation agreements;

3. The internal situation in the country of final destination, as a function of the existence of tensions or armed conflicts;

5. The national security of the member states, as well as that of friendly and allied countries,

6. The behavior of the buyer country with regard to the international community, as regards in particular to its attitude to terrorism, the nature of its alliances, and respect for international law;

7. The existence of a risk that the equipment will be diverted within the buyer country or re-exported under undesirable conditions; and

8. The compatibility of the arms exports with the technical and economic capacity of the recipient country.

In the second part, the Code gives operational provisions for the consultation mechanism. Specifically, it states that:

Member states circulate details of licenses refused in accordance with the Code together with an explanation of why the license has been refused.
Before granting a license, which has been denied by another EU country for an ‘essential identical transaction within the last three years,’ any member state will first consult the member state that issued the denial.
If following consultations, the member state nevertheless decides to grant a license, it will notify the EU partner country issuing the denial, giving a ‘detailed’ explanation of its reasoning. Both denials and consultations will be kept confidential.
Any export decision remains at the national discretion of each member state.

The Code’s operational provisions finally set a mechanism for a more general exchange of information and consultation. Aiming at the gradual development of a common exports doctrine,

II.3 The follow-up process

Consultation about the Code’s implementation and operation takes place in COARM, the Council’s working group on arms exports. So far, COARM has submitted two consolidated reports to the Council (1999 and 2000), reviewing the experiences of the preceding year and defining the working program of the following twelve months.

According to these reports, over the past two years, COARM has focused its work on the Code of Conduct on:

Moreover, COARM discussed a number of issues related to the Code of Conduct, such as procedures for monitoring arms brokers activities and national export control policies vis-à-vis certain embargo-free countries that are being closely monitored.

COARM’s biggest success was undoubtedly the establishment of a Common List of Military Equipment, qualified as a top priority in the first report.[15] This list is vital for the operation of the Code, because the use of divergent national control lists has been a major source of incoherence. The new common list clarifies and simplifies the information exchange among member states, allowing them to use common references in their denial notifications. In the medium term, it could even replace national export lists and serve as a basis for updating the list annexed to Article 296 (or simply replace it).

Progress was further made on the content of denial notifications. In the first year, member states made reference only to the number of the respective criteria when they informed their partners about a denial, without providing any further explanation. In order to facilitate understanding of the general thinking behind each other’s decisions, member states agreed that denial notifications should be more comprehensive and include not only the country of destination and the date of denial, but also a full description of the goods concerned (with their matching Common List number), a specification of the buyer (police, army, navy, etc.), a description of the end-use, and the reasons for denial (including not only the number(s) of the criteria, but also the elements on which the assessment is based).

In contrast, no consensus has been reached so far on a common understanding of what ‘essentially identical transaction’ means. As the criterion for triggering the consultation mechanism, this concept is, of course, a key element for the Code’s operation. Based on the new common list, it should be easier for member states to come to a common definition of ‘essentially identical transaction.’ It remains on the second report’s priority list for 2001, together with:

II.4 Assessment

The Code of Conduct was both welcomed as an important step towards a common European approach to arms exports and criticized for its numerous loopholes and shortcomings. Critics highlight the following weak points:

The fact that the Code was adopted as a Council Declaration under the CFSP rules means that it is politically, but not legally binding. This means that the Code can only be as strong as the will of the member states to respect it. Even within the CFSP framework, the Council could have made use of more restrictive instruments (Joint Action, Article 14 TEU).

Adopted within the framework of CFSP, the EU Code was subject to the constraints of unanimity, leading inevitably to vague formulations and compromises based on the smallest common denominator. For example, the commitment ‘not to issue an export license if there is a clear risk that the proposed export might be used for internal repression’ is flawed because of the word ‘internal.’ Member states could claim that this consideration only applies to serious abuses of human rights that occur within a recipient government’s borders (excluding, for example Turkish troops operating against Kurds in northern Iraq).[17]

The principles outline broad moral and political considerations that licensing authorities have to take into consideration. But there has been little progress in establishing prescriptive criteria that could help define these principles. For example, member states have formally committed themselves to considering the effect of arms sales on the economies of importing countries, but they have not adopted a common methodology to determine ‘detrimental effect’ (such as an excess of military expenditure over public health and education expenditure). The requirement to examine an importing country’s attitude to terrorism is not underpinned by any explicit criterion either (like the adherence to specific international conventions).[18]

The Code does not cover all relevant areas. It does not, for example, include any constraints on the activities of international arms brokering agents. Licensing of production abroad has not been addressed either.

If one EU country wishes to take up a license that has been denied by another, it needs to notify and consult only with the member state that first issued the denial. Limiting consultation on undercutting to bilateral exchanges shows how weak the Code is compared to the dual-use regime (see above). Reducing the peer pressure that is supposed to lead to a common export doctrine and self-restrain ‘has been the most criticized aspect of the Code. Many of the EU governments were keen to see, at the minimum, multilateral notification of an intention to undercut (as distinct from multilateral consultation before a decision is taken), fearing that if negotiations and information exchange is limited to the bilateral context a common approach would not be achievable.’[19]

Covering only denials and undercutting, the Code’s consultation mechanism is limited to a rather small part of arms exports. Beyond this and discussions of general export policy, there is no consultation or information exchange on granting specific export licenses. The impact of peer pressure on self-restrain remains, therefore, limited.

The Code underlines the importance of increased transparency, but it contains no reference to a public or parliamentary review of either the Code or the member states’ implementation of the agreement.

The Code does not provide for the harmonization of control systems. For example, there is no agreement among the 15 members on the best way of certifying end-use, and very few of them have follow-up mechanisms.

The Code of Conduct is, of course, easy to criticize, but any assessment has to take into account the complexity of European politics. The idea, for example, that the Code would be more effective if it were legally binding does not seem very realistic. First of all, the Code and its obligation for consultation was acceptable to the big arms exporting countries only because it was not legally binding, and any attempt to go beyond this obligation would have failed. Second, as long as the criteria for issuing denials remain general - and they are general even at the national level - member states will always have plenty of room for interpreting them according to their interests. As long as arms exports do not fall under Community law, the effectiveness of the Code’s principles and provisions will always depend on the member states’ political will - whether the agreement is legally binding or not.

Clearly, for some governments, demonstrating to their own public their commitment to moral principles has been more important than focusing on substance. Nevertheless, experience so far has shown that, in general, member states seem willing to respect their commitments. Denials have been issued, consultations have been taking place, and measures have been approved to improve the Code’s operation. On the other hand, it is difficult to say whether progress has been made as a result of peer group pressure among EU members or due to pressure from national public opinions on governments to pay more than lip service to an ‘ethical’ foreign policy. In this regard, the growing transparency that national authorities demonstrate in the field of arms exports is indicative. (Even though there is no provision for it in the Code, most governments, and in particular those of the big exporting countries, in fact now publish national reports on their arms export policy).

It remains, however, difficult to measure the effectiveness of the Code. The second report notes a ‘considerable increase in the number of notified denials and consultations,’ and considers this as an indicator of ‘member states’ resolve to put into practice a new form of transparency in arms export control and to act in greater concert in this area.’ The conclusion, however, that this reflects the Code’s rising impact, is less evident. First, some countries felt ‘obliged’ to notify denials in order to demonstrate their respect for the political commitment they had made. Second, the report only notes the number of denials, but not the financial or strategic importance of the export concerned. Last but not least, a comparison of sheer numbers can be misleading. In 1999, for example, France notified 62 denials and Sweden not a single one, although the latter is supposed to have a more restrictive export policy than the former. In fact, the differences come from the specificity of the respective licensing procedures: Swedish companies contact their government informally beforehand and request an export license only if they are sure to obtain it.

These problems are due to a lack not of political will but of harmonization between national export control regimes. The Code ‘only’ provides a common set of ethical standards beyond which national control regimes continue to differ widely.[20] Important differences persist at the administrative level concerning, for example, the scope and level of control lists or the departments and executives grades responsible for individual licensing decisions. Differences at the political level are even more important but harder to assess, since arms export policies must be inferred from actual administrative practice. What is evident is that arms export policies vary widely in accordance with the different foreign, economic, industrial and security interests of member states.

Whereas member states with no or minor defence industrial assets tend to argue on moral grounds, arms producing countries have to strike a balance between ethical claims and their industrial interests. France and the UK, in particular, focus less on general policy and more on the specific nature, destination, and economic benefit of individual exports. From their point of view, licensing decisions are to be determined according to the circumstances of each application; consistency with administrative case law is much more important than the adherence to abstract ethical principles. This approach does leave space for other nations’ perspectives to be taken into account, but only so long as they have a material bearing on the particular case under consideration. No allowance is made, in contrast, for the views of nations that might have concerns about, but no direct interest in, the outcome of individual licensing applications.[21] The Code of Conduct reflects this approach; it expresses a determination to set high common standards for arms exports, while disallowing member states any formal influence over the licensing decisions of other EU countries - save in the specific instance of ‘an essentially identical transaction’ within a three-year period.

Some areas are not treated in the Code, because they were considered as either too complex or too specific. Licensing, for example, is for exporting countries an issue more of protection of technology than of moral reasoning. Activities of armaments brokers are extremely difficult for EU governments to control. EU members do not have extra-territorial laws necessary for an effective control of brokers (who normally operate on an international level). In addition, armaments brokers often trade (small) weapons produced outside the EU (particularly in Eastern Europe). Member states actually prepare national regulations in this field, and probably the next Code report will highlight efforts to improve cooperation. Officials, however, doubt the efficiency of the envisaged measures, considering them as pure declaratory policy.

Nevertheless, in spite of all its shortcomings, the Code of Conduct represents the most comprehensive agreement to date in terms of multilateral efforts to specify how human rights, regional security and development concerns should be addressed within the export licensing process. It is a political signal, but it can increasingly serve as a framework for dialogue on more substantial issues as well. Regular consultation, joint assessment, and the progressive work on common definitions are the only means to come to a common understanding of the agreed principles. Any assessment of the Code has to take into account that in Europe, important divergences on arms export policies persist and that harmonization is still at a very early stage. Therefore, the COARM work and its results often look modest and unspectacular, but they are essential to create a basis for a common policy in this area. The new guidelines for formulating denial notifications and the envisaged harmonization of statistics, for example, are prerequisites for a growing convergence. The more consultation extends beyond specific cases of undercutting and into broader political issues (like the assessment of specific regions), the closer EU members will come to a common export doctrine. This development will be interrelated with the broader movement towards a true CFSP. It might be a time-consuming and awkward process, but one without real alternative under current circumstances.

CHAPTER THREE: FRAMEWORK AGREEMENT CONCERNING MEASURES TO FACILITATE THE RESTRUCTURING AND OPERATION OF THE EUROPEAN DEFENCE INDUSTRY

In July 1998, the defence ministers of the six major arms producing countries in Europe (France, Germany, Italy, Spain, Sweden, and the United Kingdom) signed a Letter of Intent (LoI) aimed at facilitating cross-border restructuring of their defence industries. Six ad hoc working groups were set up to establish a catalogue of measures in the following areas: security of supply; security of information; Research and Technology; harmonization of military requirements; treatment of technical information; and export procedures.[22] The working groups presented their reports in July 1999, and on the basis of their findings, an executive committee produced a final document that was signed by ministers in July 2000 as the Framework Agreement.

In contrast to the Code of Conduct, the Framework Agreement is a legally binding international treaty. It stands outside the EU context and concerns only the six major European arms producing countries. This exclusivity is based on the assumption that it would be easier to make progress, if participation is restricted to countries with a sufficient degree of common interests. The agreement is an attempt to adapt main elements of national procurement policies and regulatory frameworks to an industrial landscape that is becoming increasingly transnational. More specifically, it aims at harmonizing defence-related rules and regulations with the overall objective of creating the basis for a homogeneous defence economic space. The latter is in fact considered as indispensable to strengthening the position both of national governments vis-à-vis the new Transnational Defence Companies (TDCs) and of the latter vis-à-vis their (American) competitors.[23]

In this context, transfers and export procedures are essential. Due to developments on both the demand side (shrinking budgets, growing need for interoperability) and the supply side (cross-border consolidation), more and more weapon systems in Europe will be developed in international cooperation - either by several companies from different countries or by a single company with production sites in various countries. Such transnational programs can be intergovernmental (with several national governments as customers) or purely industrial (with companies developing defence systems on their own initiative, in particular for export markets). This growing internationalization clashes with the persistence of purely national export control regimes. Companies working on a cooperative project are obliged to go through several complex export procedures when they move components and subsystems between sites located in different countries. Transit and custom requirements cause additional delays. Moreover, the awarding of an export license for a component depends on the final destination of the system: if it is a third country, the authorities of the country exporting the component may refuse delivery for political reasons. For defence companies, this possibility is an element of insecurity that hampers cross-border cooperation. The lack of harmonization in this field is a major handicap for European industries, but it is also a problem for governments. In fact, TDCs can - at least, in certain cases - move production from one country to another, choosing the most liberal export policy.[24]

III.1 Main features

The Framework Agreement tries to tackle these issues, facilitating the movement of defence goods and services among the six so called ‘LoI nations’ and defining common export procedures for cooperative projects. It covers exports and transfers in the framework of both intergovernmental and purely industrial cooperation (Part 3, Articles 12-18).[25]

Concerning ‘transfers’ within the framework of an intergovernmental program, the six nations agreed to use Global Project Licenses (GPLs) as the necessary authorization, when the transfer is (a) needed to achieve the program, or (b) intended for national military use by one of the Parties.

Such a GPL has the effect of removing the need for specific authorizations to the destinations permitted. The conditions for granting, withdrawing or canceling a GPL will be determined by each Party, ‘taking into consideration their obligations under the present agreement’ (Article 12).

Concerning ‘exports’ of defence articles produced within the framework of an intergovernmental program to third countries, Article 13 stipulates that the Parties shall agree, for each program, on basic principles governing exports and export procedures. They should set out, on the basis of consensus,

Once an agreement has been reached on these principles, the responsibility for issuing an export license for the permitted destinations lies with the Party within whose jurisdiction the export contract falls (Article 13.3).

Establishment and later additions of permitted destinations are the responsibility of those LoI countries that participate in the cooperative program. They should take these decisions by consensus after consultations, taking into account the respective national export control policies, the fulfillment of their international commitments (including the Code of Conduct criteria), and the protection of the partners’ defence interests, ‘including the preservation of a strong and competitive European defence industrial base.’ If industry desires to add later a permitted destination, it should, as early as possible, raise this issue with relevant Parties (Article 13.2 a).

A permitted export destination can be removed only if there are significant changes in its internal situation, ‘for example full scale civil war or a serious deterioration of the human rights situation, or if its behaviour became a threat to regional or international peace, security and stability, for example as a result of aggression or threat of an aggression against other nations.’ If consensus on the removal of a permitted destination cannot be reached at the working level, the issue will be referred to the Defence Ministers of those LoI countries that participate in the project. The consultation process should not exceed three months. During that time, any participating Party may request a moratorium on exports to the destination in question. At the end of that period, the particular destination shall be removed from the permitted destinations unless consensus has been reached on its retention (Article 13.2 b).

Parties not participating in the program commit themselves to obtaining approval from the other participants before authorizing re-exports to non-LoI countries (Article 13.4). Parties shall also undertake to obtain assurances from the end-user for exports to permitted destinations, and consult with the relevant Parties if a re-export request is received (Article 13.5).

When TDCs or other defence companies carry out a program on the territory of two or more Parties, they can ask their relevant national authorities to issue a GPL for that program (Article 14). According to Article 15, GPLs can also be issued at an early stage of development of an industrial cooperation for transfers concerning the exclusive use by the industries involved.

Outside the framework of an intergovernmental or an approved industrial cooperation, the Parties commit themselves to applying simplified licensing procedures for transfers of components or subsystems produced under subcontract between industries (Article 17). Moreover, they have agreed to minimize the use of governmentally issued end-user certificates and of international import certificate requirements on transfers of components in favor of, where possible, company certificates of use (Article 16).

III.2 Assessment

The Framework Agreement outlines general provisions that are now worked out in detail within specific rules and regulations. The above-mentioned provisions on transfers and exports are, of course, difficult to assess as long as it is unclear how they will be implemented and how the procedures will work in practice. In any case, it is clear that the basic idea is not to create a common defence market for all defence goods. In fact, nationally produced items are only briefly mentioned with regard to security of supply.[26] The agreement ‘only’ strives to tackle specific problems of cooperative projects.

Here, the GPL is the key element. Taking the German Sammelausfuhrgenehmigung as a model, the GPL covers all cross-border transfers between cooperating companies and their suppliers. This means that, within a cooperative program, individual controls of each transfer will be replaced by only two general controls, one a priori (at the moment when the GPL is requested) and the second a posteriori (through an audit at the end of the duration of the license).[27] This new approach streamlines procedures and facilitates industrial cooperation considerably.[28]

For transnational companies, in particular, it is important that cooperation outside intergovernmental programs can also qualify for a GPL. The limitation to industrial agreements approved by governments is rather theoretical, since all cross-border industrial cooperation has de facto government approval (transfers between subsidiaries of transnational consortia, for example, are covered by conventions that are approved by governments).

On the other hand, European TDCs are almost exclusively in the aerospace and defence electronic sectors, which are strongly export-oriented. Since these companies realize an important part of their turnover in third markets, the importance of the GPL for them will depend very much on how the provisions on permitted export destinations will be implemented.

Concerning exports of cooperatively produced systems, the agreement makes a distinction between the six LoI nations and other countries. Sales of systems jointly produced by some LoI countries to other LoI partners, who have not participated in that program, are no longer considered as ‘exports’ but as ‘transfers.’ This means that GPLs create de facto licence free zones between the six LoI nations for cooperative projects; once a GPL is issued, companies are authorized to sell a given system to all LoI countries without any need for further export licenses.

Concerning exports of cooperatively produced systems to non-LoI countries, the impact of the Framework Agreement depends on how the concept of permitted destinations will be implemented. What, for example, will be the role of industry in establishing these destinations? The agreement only says that companies can ask for ‘later addition’ of a permitted destination, whereas industry wishes to participate as early as possible in the process through a ‘right of first proposal’ for permitted destinations that would then be discussed with governments. This raises further questions: At what stage of the program will permitted destinations be defined, given that in certain cases the question of exports arises only 15 to 20 years after the project has started? Will all participating countries have equal rights in the establishment or removal of permitted destinations, or will the weight of a country’s vote depend on the importance of its industrial contribution?

These modalities have not yet been figured out, but the six LoI countries seem to have adopted a rather flexible and pragmatic approach. First, there is a tendency, at least in big exporting nations, to decouple the granting of a GPL from the establishment of permitted export destinations. In this case, the latter would not be a prerequisite for, but a simple extension of the GPL. Awkward and time-consuming discussions on exports to third markets would not delay the granting of a GPL, and industrial cooperation could rapidly benefit from streamlined procedures. Second, the initiative for proposing permitted destinations will probably be left to the companies in charge of the program. Although governments will, of course, have the final say, this approach seems to indicate that the six nations are willing to give commercial aspects a considerable weight in their export decisions.[29]

Right from the beginning, it was clear that the LoI partners would not use ‘black lists’ of prohibited destinations. Europeans traditionally reject black lists, because they want to avoid offending certain countries, but also because it is extremely difficult for them to reach a consensus on a priori prohibited destinations other than those officially under embargo. More surprising is the fact that the Framework Agreement avoids even using the term ‘list’ for permitted destinations. The concept of this ‘white list’ was hotly debated beforehand and severely criticized by industry for being too restrictive and exclusive. Each GPL will now probably have in its Annex a list of a priori permitted destinations. Moreover, there will be an ‘exportability list,’ based on proposals from the companies involved and tailored individually to each version of a given program.

Although it is not explicitly mentioned in the Framework Agreement, there seems to be a common understanding among the six partners that these lists will remain secret. In particular for Sweden, whose parliament used to be informed beforehand about important exports decisions, this could mean a considerable loss of transparency.[30] An open question remains the degree to which all LoI countries will be involved in the establishment of permitted export destinations. It was suggested that a ‘coordination group’ should be created for each program, whereby non-participants would be consulted on decisions taken by participating nations. This approach could indeed contribute to the development of a truly common export policy. On the other hand, it would raise serious problems for the protection of commercially sensitive information.[31] Therefore, it is more likely that non-participating LoI countries will, at best, be informed only after a decision concerning a particular program has been taken.

Provisions on exports are a compromise between advocates of a liberal export policy - namely Great Britain and France - and those countries that have traditionally held a more restrictive stance - in particular Germany. For the three most important arms producing and exporting countries, the Framework Agreement is all the more important since it replaces de facto the so-called Schmidt-Debré accord that had governed Franco-German cooperative programs ever since the early 1970s. (British-German projects, in particular Tornado and Eurofighter, have been managed under the same principles). The central element of the Schmidt-Debré accord was a tacit agreement that allowed the country holding the export contract to take the final decision. In practice, this means that Germany (as the country with the more restrictive export policy) abstained from its veto right on French (or British) authorizations to export jointly produced systems. During the Cold War, Germany accepted this ‘gentlemen’s agreement’ not only because of the latter’s commercial benefits, but also because of its own political weakness vis-à-vis the two major European powers. German unification put an end to this political imbalance and shattered the basis of the Schmidt-Debré agreement. When the newly elected red-green government refused authorization for the Tiger helicopter to take part in flight demonstrations in Turkey in December 1998, it became clear that Berlin would no longer give carte blanche for exports of jointly produced systems.

The Framework Agreement takes these developments into account; it confirms each participant’s formal veto right on exports, while allowing room for flexible interpretation. Flexibility will probably characterize the implementation of the agreement (see above), but it can also be derived from its wording. The agreement outlines, for example, that decisions on the establishment of permitted export destinations ‘should’ be taken by consensus (instead of the initially envisaged ‘shall’), which seems to weaken the individual veto right. According to the agreement, a permitted destination will have to be removed, if one participant asks for it. On the other hand, the agreement is very detailed on the conditions and the procedures for such a removal (in very specific cases and after full consultation only). Again, the theoretically strong position of each participant seems to be counterbalanced by rather high implicit barriers against the actual right to ask for removal (which is, of course, important for the financial stability of a program).

In this context, the definition of ‘participant’ will be important as well. It seems clear that, within an intergovernmental program, each participating country that acquires the system will be considered as a participant (and will therefore participate fully in the establishment of permitted export destinations). Within a purely industrial cooperation framework, there will probably be a threshold (based on the financial value of the respective industrial share) to distinguish between participating firms and suppliers. Only countries hosting a participating firm will have a say in the establishment of permitted destinations. Another question concerns projects in which non-LoI countries participate (like, for example, the Airbus A400 M with 5 LoI nations plus Turkey, Belgium, and Luxembourg). Will it be possible to use GPLs for these projects, and if so, how? LoI nations have made it clear that they would prefer ad hoc arrangements in these cases rather than accepting/ inviting other countries to become full parties to the Framework Agreement. What these ad hoc arrangements will look like, however, is unclear.[32]

The introduction of a Global Project License can be considered as a sign of political will to move towards a common export policy. On the other hand, the Framework Agreement in itself provides for nothing more (and nothing less) than simply a mechanism for intergovernmental consultation and decision-making. Concerning the substance of their export policies, there are still important divergences between the six partners. Granted, there is consensus among them on prohibited destinations that are covered by embargo lists (although there are still diverging interpretations of definitions). At the other end of the spectrum, there is also consensus on certain permitted destinations, in particular other EU members and NATO countries (with the exception of Turkey). But, even in these cases, it is not excluded that LoI partners may disagree on which version of an equipment is to be exported.[33] The real problems, however, will arise when it comes to ad hoc decisions on exports to certain regions like the Persian Gulf. Whatever the details of the export procedures in their final state, the real issue will remain a political one. As long as European consensus concerns only general principles but not their practical interpretation, the traditional disagreements will probably reappear when it comes to concrete decisions.[34]

The coexistence of three differently organized systems shows the lack of coherence that characterizes Europe today in the field of strategic exports. Whereas the dual-use regime is part of EU pillar one (EC), the Code is established under pillar two (CFSP). The third element, the Framework Agreement, remains completely beyond the EU framework.

The Community export control regime for dual-use goods is based on mutual recognition of national export decisions. Since the reform of the system in June 2000, it provides for intensive consultations on the functioning of the system. Through the Annexes and the explicit references to multilateral non-proliferation regimes, the Regulation makes the provisions of the MTCR, Wassenaar, and the NSG legally binding for all EU member states.[35]

In spite of all its shortcomings, the dual-use regime is a much stronger instrument than the Code of Conduct for Arms Exports. The Code only provides for broad common standards; it is not legally binding and stipulates specific consultation only in certain limited cases.

Finally, the Framework Agreement between the six major arms producing EU countries sets up an export regime for cooperative armaments projects. It establishes a de facto license-free zone among the six participating countries and a decision-making mechanism for exports of jointly developed weapon systems to third countries. In contrast to the dual-use regime, export provisions of the Framework Agreement operate on an ad hoc basis, depending on the participating countries and the respective cooperative program. Free movement of goods (and services) is possible only if a GPL is granted and remains limited to specific cooperative programs. Export authorizations to third countries are recognized only if there is agreement on permitted destinations.

One thing all three systems have in common is that they do not constitute a common (export) policy. On the contrary, whilst the dual-use regime is built upon the mutual acceptance of divergent policies (reflected in mutual acceptance of export authorizations), the Code of Conduct merely tries to limit the worst excesses of political divergence on arms exports and the Framework Agreement introduces a mechanism that allows only ad hoc agreements. At the same time, all three systems are supposed to foster convergence among participating countries and thus lead, eventually, to a common policy.

The main instruments to promote progressive harmonization are cooperation and consultation. With regard to dual-use exports, the new provision for consultation on undercutting should help harmonize national policies, but its scope (and therefore its political impact) could be limited by the increasing liberalization of dual-use trade. Moreover, consultation within the LoI framework will be highly political and sensitive. Whereas consultation on dual-use exports is politically toned down through mutual recognition of export authorizations and the existence of a white list of destinations for (almost) all items, the LoI nations will need to secure consensus on export destinations for each particular program, in varying formats. This means that the LoI consultations will be more substantial than those under the Code of Conduct. They have a different (or complementary) philosophy (focusing on permissions rather than refusals), involve the biggest arms producers, cover the most important programs, and are much more detailed.

At the very core of export decisions, the LoI consultation process could become a powerful and effective driving force towards European convergence. This does not mean that a rapprochement will be easy to achieve. Indeed, conflicts between the current German government and its French and British counterparts seem preordained. Nevertheless, the ongoing cross-border restructuring of European defence industries is putting all LoI governments under pressure to harmonize their export policies. The most ‘Europeanized’ sectors, namely aerospace and defence electronics, are actively lobbying for a broad application of the principles of the Schmidt-Debré accord within the context of the LoI process, leaving the final decision to the country that holds the contract.[36] In accordance with the principles of the dual-use regime, such an agreement would allow the free circulation of defence items among the parties concerned. From an economic point of view, this would indeed be the most rational and logical solution. In the foreseeable future, however, such an agreement could, at best, only be reached outside the EU framework among the LoI partners.

* Burkard Schmitt is currently a research fellow at the Institute for Security Studies of WEU. He was formerly an independent researcher and journalist, writing in particular on the Franco-German relationship, the French Fifth Republic and nuclear deterrence. At the Institute for Security Studies he covers nuclear issues, defence industry and armaments cooperation, organising a defence industries task force. Institute publications include Chaillot Papers 44, ‘Between cooperation and competition: the transatlantic defence market’ (ed.) and 40, ‘From cooperation to integration: defence and aerospace industries in Europe’, and Occasional Papers 3, ‘L’Europe et la dissuasion nucléaire’ and 18, ‘National missile defence and the future of nuclear policy’ (joint ed.).

[1] The scope of Article 296 is defined by a rather restrictive list of arms and munitions that the Council drew up in 1958. The list is exhaustive, in the sense that products not included in the list are not covered by Article 296. Since it has not been changed since 1958, Article 296 does not apply to many modern strategic goods and technologies.

[2] Council Regulation (EC) n° 3381/94 and Council Decision 94/942/CFSP are published in the Official Journal of the European Communities, L 367, 31 December 1994, pp. 1–7 and pp. 8-163.

[3] Under Article 133, paragraph 1, ‘The common commercial policy shall be based on uniform principles, particularly in regard to changes in tariff rates, the conclusion of tariff and trade agreements, the achievement of uniformity in measures of liberalisation, export policy and measures to protect trade such as those to be taken in the event of dumping or subsidies.’

[4] A Regulation is legally binding in all its elements. There are Council and Commission Regulations, but their nature and effects are identical. Becoming part of Community law, Regulations are directly applicable in each member state; transposition into national law is automatic. The Regulation comes into force simultaneously and uniformly in each member state. By contrast, a Council Decision on a joint action is politically, not legally binding. Joint actions are instruments of the Union’s CFSP (pillar II). According to Article 14 of the Treaty on European Union (TEU), joint actions ‘shall address specific situations where operational action by the Union is deemed to be required. They shall lay down their objectives, scope, the means to be available to the Union, if necessary their duration, and the conditions for their implementation.’

[5] According to Annex III, member states would take into account commitments they have accepted in other non-proliferation arrangements, obligations under sanctions imposed by the UN or the EU, considerations of national foreign and security policy, intended end-use, and the risk of diversion.

[6] In its initial version, Annex II included Australia, Austria, Canada, Finland, Japan, Norway, Sweden, Switzerland and the United States. Austria, Finland, and Sweden were removed from the list when they became themselves EU members. Later on, three EU candidates and New Zealand were included. See also footnote 11.

[7] The European Atomic Energy Community (Euratom) Treaty, one of the three treaties establishing the European Communities in 1957, is part of the EU’s first pillar.

[8] Council Regulation (EC) 458/2001 of 6 March 2001 amended Regulation (EC) 1334/2000 with regard to the export list of controlled dual-use items and technology in order to reflect the liberalization of control parameters agreed on 1 December 2000 among the participating states in the Wassenaar Arrangement. Council Regulation (EC) 1334/2000 and Council Decision 2000/402/CFSP are published in the Official Journal of the European Communities, L 159, 30 June 2000, pp. 1-218. Council Regulation (EC) 458/2001 can be found in the Official Journal, L 65, 7 March 2001, p. 19.

[9] According to the ECJ, an integrated system is permitted only if Community law alone cannot provide a sufficient legal basis. In the case of dual-use goods, the integrated system was therefore a violation of Community law. See ECJ, Case C 83/94 Leifer Judgement of 17.10.95. For a detailed analysis of the legal aspects of the dual-use export regime, see Nicholas Emiliou, ‘Strategic Export Controls, National Security and the Common Commercial Policy,’ in European Foreign Affairs Review 1 (1996), pp. 55–78; and Simone Bermbach, ‘Die gemeinschaftliche Ausfuhrkontrolle für Dual-use-Güter’ (Baden-Baden: Nomos, 1997).

[10] Report to the European Parliament and the Council on the application of regulation (EC) 3381/94 setting up a Community system of export controls regarding dual-use goods, Com (98) 258, Brussels, May 1998, p. 10.

[11] Article 4, paragraph 2 states that ‘For the purpose of this paragraph, ‘military end-use’ shall mean: a) incorporation into military items listed in the military lists of member states; b) use of production-, test- or analytical equipment and components therefor, for the development, production or maintenance of military items listed in the above-mentioned list; c) use of any unfinished products in a plant for the production of military items in the above-mentioned list.’

[12] Australia, Canada, Czech Republic, Hungary, Japan, New Zealand, Norway, Poland, Switzerland, and the United States. The Czech Republic, Hungary, and Poland will be removed once they become EU members.

[13] At the Community level, a coordinating group, composed by national licensing officials and customs officers and chaired by the Commission, is regularly discussing the practical application of the regulation. The group focuses on resolving practical problems and developing common interpretations of certain provisions of the Regulation.

[14] Since its adoption, the Code’s principles have been recognized by the associated countries of Central and Eastern Europe, Cyprus, the EFTA countries, members of the European Economic Area, Canada, and most recently, Turkey and Malta.

[15] The list was adopted by the Council on 13 June 2000 and published in the Official Journal of 8 July 2000. Having an evolutionary character, it will be regularly updated by member states through the COARM working group.

[16] COARM has undertaken to draw up a list that will be submitted to the Commission. The latter will then be responsible for taking the initiative of proposing a draft Community mechanism for controlling these kinds of exports. Monitored by Community rules, this instrument will be separate from the operative provisions of the Code of Conduct.

[17] Elisabeth Clegg and Alexandra McKenzie, ‘Developing a Common Approach? The EU Code of Conduct on Arms Exports,’ in Bulletin of Arms Control, No. 32 (1998), pp. 22-28.

[18] Alan Piggot, ‘Western European Defence Exports: Prospects for a Common Market,’ in Defence Analysis Vol. 15, No. 2 (1999), pp. 167-184.

[22] In the Framework Agreement, the areas are organized in a slightly different way: Security of Supply; Transfer and Export Procedures; Security of Classified Information; Defence-Related Research and Technology; Treatment of Technical Information; Harmonization of Military Requirements; and Protection of Commercially Sensitive Information.

[23] For an in-depth analysis of the framework agreement, see Burkard Schmitt, ‘From cooperation to integration: defence and aerospace industries in Europe,’ Chaillot Paper 40 (Paris: Institute for Security Studies of WEU, July 2000).

[24] See Jean-Claude Sandrier, Christian Martin and Alain Veyret, ‘Le contrôle des exportations d’armement,’ Rapport d’Information de la Commission de la Défense, n° 2334 (Paris : Assemblé Nationale, 2000).

[25] According to Part 1, Article 2 of the Agreement, ‘transfer’ means any movement of defence articles or defence services among the parties, ‘export’ means any movement of defence articles or defence services from a party to a non-party.

[26] Concerning transfers between parties of defence articles and related defence services that are nationally produced, ‘Parties shall make their best efforts to streamline national licensing procedures’ (Article 17).

[27] The treaty does not indicate the duration of a GPL, but there seems to be an agreement now that it should be valid for five years.

[28] The LoI partners decided also to review, on a case-by-case basis, existing cooperative projects with a view to applying GPLs, where possible, to these programs. This is all the more important since there are very few new programs planned in Europe for the foreseeable future.

[29] Another indicator is the fact that the Framework Agreement underlines twice (in the preamble and in Article 13.2a) the importance of a strong European defence industrial base. A proposal from France and the UK to include a similar formulation in the Code of Conduct failed because of the resistance of some smaller EU countries.

[30] See Michael Brzoska and Hartmut Küchle, ‘Folgen, Auswirkungen und Gestaltungsmöglichkeiten internationaler Abkommen für eine restriktive deutsche Rüstungspolitik,’ Gutachten für den Wissentschaftlichen Dienst beim Deutschen Bundestag, Bonn International Center for Conversion, January 2001.

[31] If decisions on permitted destinations are based on proposals from industry, such a consultation mechanism would allow all members of this coordination group to be informed about the commercial strategies of the companies concerned. The risk is that this information could then be ‘leaked’ to potential competitors in other (non-participating) LoI countries.

[32] Conditions for signing up to the framework agreement are quite demanding and discriminatory. Other EU members can apply for membership, in which case the six examine the candidature and must agree unanimously. In the case of European non-EU members, an invitation to join must come from the six who, once again, must all agree on this.

[33] The only zone where there is no restriction at all (including the willingness to export the most advanced version of a system) is that of the six LoI Parties themselves (forming what one official called a ‘virgin list’).

[34] Persisting political divergences among the six LoI nations are the main reason for the industry’s lack of enthusiasm about the agreement. Without ignoring the potential benefits of the LoI procedures, many industrialists fear that intergovernmental decision-making based on consensus will reduce excessively the number of permitted export destinations. French and British companies, in particular, are worried that German reluctance could make an LoI policy too restrictive (compared to the national policies of their own governments). There is also the fear that, on certain occasions, some governments might be put under U.S. pressure to use their veto against an export decision. On the other hand, the envisaged decision-making mechanism can provide for considerable peer pressure on reluctant partners.

[35] The recent Council Regulation (EC) 458/2001 of 6 March 2001 amending Regulation (EC) 1334/2000 was explicitly adopted ‘in order to enable the Member States and the Community to comply with their international commitments’ and, more specifically, with the changes in the control parameters of dual-use items and technology agreed at the sixth plenary meeting of the Wassenaar Arrangement. (See also comment in footnote 8.)

[36] In this context, it should not be forgotten that not all defence industrial sectors are equally transnational. Naval shipbuilders and land armaments producers in Europe are still mainly national and compete with each other in markets all around the globe. With demand in home markets flat, exports are vital for these industries, and governments are, for social and economic reasons, normally very active in supporting them abroad. This intra-European competition in third markets is a strong argument for European governments to maintain their own national export policy, and it is unlikely that this situation will change in the foreseeable future. This does not mean, however, that the free movement of goods would not be possible (as the dual-use regime has shown).

A COMMON EUROPEAN EXPORT POLICY FOR DEFENCE AND DUAL-USE ITEMS?

CHAPTER ONE: THE COMMON CONTROL REGIME FOR DUAL-USE EXPORTS

I.1 The 1995 control regime

I.2 The review process

I.3 The 2000 control regime

I.4 Assessment

CHAPTER TWO: THE CODE OF CONDUCT

II.1 Export criteria

II.3 The follow-up process

II.4 Assessment

III.1 Main features

III.2 Assessment