Introduction

Anyone familiar with the common-law system might be surprised when reading the case law of the European Court of Justice ECJ by the absence of any attempt to disguise the creative interpretative role of the judges. An examination of the decisions relating to the concept of direct effect (particularly of directives), a topic Invariably included in a European law course, illustrates this.

Despite the candour with which the significance of judicial decision-making and the inevitability of interpretation is conceded in the context of the European Community (EC), the role of the ECJ is still presented as a politically neutral one. The choices made by the judges of the ECJ are Important but they are not apparently political. It is interesting to ask, very briefly how this picture of non-political creativity is sustained.

It is made possible by the notion that the judges or their concerns are somehow above politics. The judges are said to favour teleological interpretation - that interpretation which best promotes 'European integration' or 'an ever closer union between the peoples of Europe' (preamble to the Treaty of Rome). Far from taking political decisions, the Judges supposedly try to implement a vision of Europe which one can infer is shared by the member states given their accession to the EC. The judges therefore seek to rise above the petty prejudices and anxieties motivated by crude nationalist sentiment of particular governments at particular moments. It is opposition to rather than the promotion of this goal of' European integration' which should be labelled as political.

In this short contribution I want to suggest that the political implications of this teleological approach are indeed great. In particular i will argue that this approach is significant as it leads to the prioritisation of certain areas of community law and the marginalisation of others. I will illustrate this with reference to the role of international treaties in EC law.

Priorities and Margins in EC Law

Since the emergence of the Commission White Paper in 1985 on the completion of the internal market, the concept of European integration seems largely to be understood through reference to the 1991 programme of attempts to secure an internal or single market in Europe. These political priorities, earlier discussed in terms of a common market, have quickly become legal priorities. As a result many legal norms which apparently form an integral part of community law have been condemned to the margins of the system, unenforceable and largely impotent. One way in which this has occurred is through the application of the legal concept mentioned above, that of direct effect.

This can well be illustrated by reference to international treaties within the community legal order. The EC is a significant world actor. It enters Into a myriad of agreements with non-member states and groups of states. With 19 per cent of world trade it is not surprising that many of these concern matters of trade, in particular issues of access to the EC market for goods originating from outside the community. The rules contained in such treaties form an 'integral part of community law'. As such the GATT code containing norms regulating the inter- national trading system constitutes a part of community law (Petersmann, 1983a). Moreover, formally these norms take precedence over both conflicting provisions of community law and the national law of the member state.

This statement of the legal position of the liberal principles of the GATT in community law will be striking to those who know even a little about the operation of the community's Common Agricultural Policy (CAP). The protectionist nature of this policy is well known and there can be no doubt that it operates in contravention of the GATT code.1 The CAP continues to absorb more than half of the total budget of the EC. Moreover, in an attempt to mitigate the effects of over-production in the community, more than one third of this sum is committed to the payment of export subsidies to farmers seeking to sell on the world market.

The negative impact of this policy on the world's poorest countries is clear. It has contributed to the dramatic reduction in world prices for certain commodities over the last two decades. Even for those countries not competing directly with EC producers the results can be disastrous. Cheap EC food imports can be politically useful for many governments of poor countries. However, these imports tend to undermine the viability of domestic initiatives in food production and thus the possibility of food security in the long term.

It is perhaps less well known that much of the legislation which constitutes the CAP is, according to the GATT and, therefore, according to EC law itself, liable to condemnation. In legal terms and in terms of priorities and margins it is worth inquiring how this situation can exist and persist.

For the answer we must begin by looking more closely at the concept of direct effect. While the GATT, as noted above, forms an 'integral part of community law', it does not form a directly effective part (Intentional Fruit Company [1972] ECR 1227). This means that the norms contained in the GATT are not capable of being invoked by individuals before national courts. Moreover a national court is not able to request a ruling, by means of Article 177, from the ECJ in order to question the validity of community law in the light of the GATT. In fact the only route open to individuals seeking to challenge community law on this basis is in a direct action before the ECJ, relying primarily on Article 173 (Hartley, 1981, Chs. 11 and 12). The difficulties involved in bringing such an action are notorious. Thus, while the EC has been careful to confer upon its institutions the means to retaliate against its trading partners who do not honour international law relating to trade,2 it has been less careful in ensuring its own conformity. It is one thing to declare a theoretical commitment to certain principles, thus gaining political kudos; it is quite another to ensure respect. It is clear then that all community laws are not equal since non-directly effective treaties remain on the margins and are therefore more often honoured in the breach.

The marginalisation of certain areas of community law as a result of a refusal on the part of the ECJ to attribute direct effect forces us to ask questions regarding the criteria invoked in conceding or denying this quality. The question of direct effect has been defined as a question  of interpretation and thus a matter to be determined by the ECJ. This court has evolved a test to determine whether a particular provision of EC law is capable of employing direct effect. Hartley, in his standard text book on the institutions of the EC, extrapolating from the case law, considers that the following conditions must be satisfied (1981, p. 188): the provision must be clear and unambiguous; it must be unconditional; and its operation must not be dependent on further action being taken by the community institutions or national authorities.

The problems arising are well illustrated by the recent case of (Demirel v. Stadt Schwabisch Cmnd ([1989] 1 CMLR 412), which suggests that the accordance of the quality of direct effect is not merely a matter of the objective application of the above criteria but also a question of political priority. Ehlermann, the director-general of the Legal Service of the Commission of the EC, admits that the value of the rule in question is a relevant consideration in determining whether it should be directly effective.

The Demirel case concerned a Turkish woman who married a Turkish national who was working in West Germany. Mrs Demirel sought to establish her right to remain with her husband in West Germany. This tight was denied on the basis that German law, since 1982, requires a worker to live for eight uninterrupted years in the state before acquiring any right to be joined by his or her family.

Mrs Demirel subsequently sought to rely on EC law, which, given the principle of the supremacy of community law, prevails over inconsistent national law, "The relevant international law was in the form of a treaty between the EC, the member states and Turkey. The Ankara Agreement was signed In 1963, its Additional Protocol in 1970. The agreements were concluded in a period of mass labour migration to Europe, and shortly after the building of the Berlin wall stemmed the flow of labour from Eastern Europe. During the post war economic boom years, this migration was encouraged by the governments of the Western European states, as is evidenced by the setting up of national immigration offices. The Ankara Agreement sought (Inter alia) to secure the flow of workers from Turkey by guaranteeing certain minimum rights for the migrants. Article 12 of the agreement notes that' the contracting parties agree to be guided by Articles 48, 49 and 50 of the Treaty establishing the Community for the purpose of progressively securing freedom of movement of workers between them'. Thus it could be expected that Turkish workers would enjoy rights akin to those migrating in an intra-community context, including the right to migrate, to reside, to remain; the right to equal social and tax advantages, and the right to family reunification. These rights were to be secured 'by progressive stages' by December 1986.3

However, given the changing labour requirements of European industry during the 1970s and 1980s, the legislative measures to ensure the progressive implementation of these rights failed to emerge. Despite this Mrs Demirel persisted. She sought to establish that Article 12 and Article 36 should be considered directly effective provisions of community law which could be relied on in national courts by individuals seeking to challenge inconsistent national law. Faced with this question (a question of interpretation) the Administrative Court  Baden-Wurttemberg made a reference for a preliminary ruling from the ECJ.

That the ECJ denied the direct effect of these provisions is hardly surprising in the light of the political interests of Western Europe in the late 1980s. However, that this decision is surprising in legal terms becomes apparent when one considers the attitude of the ECJ to the direct effect of 'internal' (that is, not derived from treaties between the EC and non member states) community law. The two cases briefly discussed below should suffice to illustrate this.

Early in the history of the EC, the ECJ determined that Articles 48 to 50 were concerned with the free movement of persons in the community and were capable of conferring rights and obligations on individuals which were enforceable in the national courts (Van Duyn v. Home Office [1975] 1 CMLR 1). Given the parallel nature of Article 12 to Articles 48 to 50, this  undermines the argument of the ECJ in Demirel and exposes its political reluctance to concede direct effect.

The second case, Defrenna v. SABENA ([1976] 2CMLR 98), concerned the possibility of Article 119 being directly effective. This article states: 'Each Member State shall during the first stage ensure and subsequently maintain the application of the principle that men and women should receive equal pay for equal work.'

It is clear that this provision required action on the part of the member states to bring the principle into operation. Nonetheless the ECJ concluded In its judgement that given the expiry of the deadline the article was itself capable of direct effect. Given the parallel nature of this article with Article 36 of the Additional Protocol to the Ankara Agreement, Hartley's comment [1981] is worth bearing in mind. He notes that this modification of the original rule to a large extent nullifies it, since almost all community provisions requiring further action contain a time-limit. In such cases the only consequence of the requirement is that direct effect is postponed until the deadline has passed.

Given the case law of the ECJ as exemplified by these two cases, and the obvious attempts of the ECJ to stretch the range of application of the concept of direct effect to its very limits in the interests of the effetutile of community law, the legal arguments in Demirel, which on their face seem convincing, appear less satisfactory. One might reasonably assume on the basis of such past decisions that the relevant articles of the agreement between Turkey and the EC are capable of direct effect. That this was denied can perhaps best be attributed to the teleological nature of the court's approach to interpretation (discussed above). This results in the court emphasising the context and more particularly the objectives of the provision in question. One result of this approach is that it allows the court to interpret identical words differently if they are placed in different contexts. Thus, in (Polydor v. Harlequin Record Shops ([1982] 1 CMLR 677), a provision of the Free Trade Agreement between Portugal and the community was held to be different in its scope from Article 45 which on its surface was identical. After all, the ECJ emphasised, this agreement sought to establish a link between Portugal and the community of a less 'intense' nature than that between the member states themselves. The objectives of the rules, as inferred from their context, could be distinguished.

Thus it is easier to understand the decision of the court in Demirel not in terms of an objectively applied legal test, although the ECJ did  purport to apply these tests, but in terms of its political context and the changing objectives of the Association Agreement as understood from a Western European viewpoint. By the 1970s immigration had been defined as a problem in Europe, a problem frequently discussed together with drug-trafficking and terrorism.4  While from a Turkish perspective access for its citizens to foreign labour markets had become of increased importance, the EC no longer welcomed these workers or their family members. Teleological interpretation allows these political needs to be translated, by means of' context' and 'objectives' or 'value', into legal realities.

The EC as an International Actor

Such a discussion gives rise to questions about the nature of international treaty negotiations. It undermines the picture of these as freely negotiated 'contracts' which lay down fixed and clearly established rights and obligations, the balance of which is mutually satisfactory to all parties. The capacity of the community to devalue the rights accruing to its legal partners, or to render these rights irrelevant in the face of political realities through legal interpretation, demonstrates that recourse to the law does not exclude notions of power in international relations.

This conclusion is perhaps hardly startling. A critic might point out that such a discussion of the interpretative tools of the ECJ is unnecessary. After all one only has to look at the text of an agreement such as the Lome Convention to be convinced of the significance of political power in the conclusion of international agreements. Lome IV is an agreement between the EC and 68 African, Caribbean and Pacific states (ACP).5 It is the showpiece of the community's development policy. It aims, according to Article 1, to 'promote and expedite the economic, cultural and social development of the ACP state ...'.

It seeks to achieve this through a judicial mix of both aid and trade. Given the relatively small scale of the aid package the trade element is of the utmost significance, particularly from an ACP perspective, insofar as it guarantees access to the EC market for ACP products. However, the degree of flexibility surrounding the EC's obligations, and its ability to tailor these to its own changing political and economic needs, is astounding. The development objective of the convention should, it seems, at no time interfere with the operation of the domestic policies of the EC. This is true even of those policies, particularly the CAP, which are in themselves antithetical to the concept of Third World development. Those agricultural products which form part of a common market organisation or regime are excepted from the general principle of free trade on which Lome is based. Moreover no definitive list of such products was established at the time of the conclusion of the convention. The EC is able to unilaterally extend the scope of this exclusion, as occurred upon Spanish and Portuguese accession to the community. One might expect a broadening of this exception in the light of recent changes in Eastern Europe and in particular the unification of Germany. For certain ACP states whose economies are largely dependent on a few commodities the effect can be catastrophic.

It is, thus, clear that the Lome Convention, like the Ankara Agreement, does not prevent the preference of EC interests. However, while this is apparent from an analysis of the text of Lome this is not true of either the GAT~ or the Ankara Agreement. When the latter was concluded Turkey enjoyed a strong bargaining position. It could offer to the West a commodity which it desperately needed - labour. It was in a position to resist the 'but if' clauses included in Lome. What is interesting then is the way in which the EC introduced the same insecurity into this apparently stable agreement. This it did through the application of a carefully contrived legal concept, apparently objectively  applied and devoid of political significance. What is clear then is that this apparently uncontentious legal tool has accorded to the ECJ the power to render worthless any law which has, in political terms, become an error of judgement.

What is the significance of this? The answer which should surely form a central part of any EC law course is that the EC is a political grouping whose politics do not stop at the door of the ECJ. Underlying, and implicit in, the court's intention to promote 'European integration' are political premises which have a significant and debilitating effect upon non-EC countries. From Turkey to the Lome countries their very ability to exist and provide for their citizens is affected in a covert and apparently non-political way. Just as the decision in Demirel  negated a fairly negotiated and apparently binding treaty, so the decisions concerning GAT and the Lome countries legitimate the maintenance of protectionist measures on the part of the EC. The operation of the CAP not only makes such commodity production in poor countries uneconomic, but also, through the dumping of excess EC subsidised produce, destroys the alternative markets that might exist. EC law is politics and the results are both crucial and devastating.

Notes

1. See Regulation 2641/84 which allows the EC to retaliate against trading partners who engage in 'illicit commercial practices' defined by reference to international law. Also Regulation 2423/88 enables the EC to impose anti-dumping duties in the event of unfair price discrimination. The operation of this law has given rise to accusations that its aim is a protectionist one.
2. See note 1, above.
3. This was laid down in Article 36 of the Additional Protocol which states 'Freedom of movement for workers between member-states of the Community and Turkey shall be secured by progressive stages in accordance with the principles set out in Article 12 of the Agreement of Association between the end of the twelfth and the twenty-second year after the entry into force of that Agreement. The Council of Association shall decide on the rules necessary to  that end.'
4. For an example of the amalgamation of these issues see Completing  the Internal Market, White Paper from the Commission, June 1985, especially paragraphs 47-56.
5. For a critical analysis of the Lome Conventions see J. Ravenhill (1985) Collective Clientelism.