I. INTRODUCTION
The 15th anniversary (1994–2009) of the Agreement on the European Economic Area (EEA)Footnote 1 seems an appropriate occasion on which to assess the EFTA Court's contribution to the Agreement (almost against all odds) having reached such a respectable age.Footnote 2
With respect to the legal effect of EEA law at the national level, there is general consensus that the Agreement has developed in a ‘supranational’ direction during the course of the past 15 years, and that this can primarily be attributed to dynamic interpretation by the EFTA Court.Footnote 3 We do, however, lack any real explanation for why the EFTA Court has preferred to take a distinctly dynamic approach. As regards the interpretation of the substantive content of the common EEA rules, it seems that not only is the EFTA Court careful to avoid EEA-specific solutions that would give the EFTA States greater political leeway than the EU Member States, but—it is argued in this article—in ‘hard cases’ it appears to lean even further towards teleological (ie integrationist) interpretation than does the Court of Justice of the European Communities (ECJ).
In parallel with the EFTA Court's dynamic development of EEA law, a form of dialogue appears to have evolved between the EFTA Court and the ECJ, in which the latter seem to regard EFTA Court case-law as a relevant source for interpreting EEA-relevant EU law. Moreover, the ECJ appears to have changed its attitude to the EEA Agreement, from one of considerable scepticism to a recognition that the goal of extending the internal market to include the EFTA States is actually achievable.
Whether there is any connection between the EFTA Court's dynamic interpretation of EEA law and the ECJ's changed perception of both the EFTA Court and the EEA Agreement is a key question. My thesis is that the EFTA Court's development of EEA law over the last 15 years can be understood as an enduring attempt to convince an initially sceptical ECJ that the EEA Agreement is a viable structure. In the following, I will argue that the EFTA Court has acknowledged that the fate of the EEA Agreement hangs on its acceptance by the ECJ. Moreover, the EFTA Court has been aware that the ECJ, in its consideration of the internal effect that an international treaty such as the EEA Agreement should be given in the Community (now EU) legal order, appears to attach importance to how effective it has become in the other Contracting Parties' legal systems. Based on the consideration of reciprocity, the development of EEA law in a ‘supranational’ and integration-friendly direction will increase the likelihood of the ECJ granting the same rights in the common market to individuals and market operators residing in the EFTA States as to operators residing in the EU. Further, the EFTA Court has been eager to dispel any suspicion the ECJ may have had as to its independence: Established and financed by the EFTA States and made up of judges appointed by their governments alone, the EFTA Court has been anxious to be percieved by the ECJ as an independent court and guarantor of the EFTA States' fulfilment of their obligations under the EEA Agreement.
In the following, I will attempt to support this thesis through an analysis of EFTA Court case-law from the beginning of 1994 up until the present (section IV). In continuation of this review, I will attempt to demonstrate that the Court's strategy seems to have caused the ECJ to change its view of the EEA Agreement (section V). Firstly, however, it is pertinent to review the seemingly impossible task the EFTA Court was set (section II), and to substantiate the claim that the ECJ was originally rather sceptical of the whole EEA structure (section III).
II. THE CHALLENGE: A HOMOGENEOUS EEA
The task the EFTA Court was charged with on its inception in 1994 can with some truth be described as a ‘Mission Impossible’.Footnote 4 After the refusal by the ECJ in 1991 to approve the original plan for a joint EEA Court,Footnote 5 the end result was two independent courts at the international level, the EFTA Court and the ECJ, interpreting the common EEA rules. As the EFTA Court itself remarked in the L'Oréal case from 2008, it is ‘an inherent consequence of such a system that from time to time the two courts may come to different conclusions in their interpretation of the rules.’Footnote 6 In order to reduce this risk to a minimum, the EFTA Court is obliged by article 6 EEA to conform with ECJ case-law prior to the date of signature of the Agreement (1 May 1992) and by article 3(2) of the Agreement between the EFTA States on the establishment of a Surveillance Authority and a Court of Justice (Surveillance and Court Agreement—SCA)Footnote 7 to pay due account to subsequent case-law. However, in cases in which no clarifying ECJ case-law exists, the EFTA Court is in reality left with identifying to the best of its ability how the ECJ would have dealt with the submitted interpretation issue.
The challenge faced by the EFTA Court is further intensified because the EEA Agreement contains no corresponding obligation on the part of the ECJ to give consideration to EFTA Court case-law. In its opinion on the original draft agreement, the ECJ stated clearly that such a duty would be incompatible with its sole competence to determine the content of Community (now EU) law.
This structural imbalance between the EFTA Court and the ECJ increases when we consider that the EEA Agreement's objective of a homogeneous EEA is not limited to uniform interpretation of the EEA rules as such, but also—and in practice much more importantly—aims to achieve legal homogeneity between the EEA rules and the ECJ's interpretation of the underlying EU law. During the EEA negotiations, the EFTA States were very much aware that in Polydor and Kupferberg the ECJ had interpreted the provisions of the free trade agreement between the then EC and the then EFTA State Portugal differently from the virtually identically worded provisions in the EC Treaty, with justification in the different aims of the two treaties.Footnote 8 For the EFTA States, it was therefore imperative to ensure that the ECJ would interpret the EEA rules in conformity with its interpretation of identically worded provisions in EU law. Concerning ECJ case-law prior to signature of the Agreement, this was to be safeguarded under article 6 EEA.Footnote 9 This provision did not, however, provide any guarantee that the ECJ would continue to interpret EEA law in accordance with subsequent case-law—guidance from the drafters of the Agreement on this matter is limited to a statement in the fifteenth paragraph of the preamble that ‘in full deference to the independence of the courts’ the objective is ‘to arrive at and maintain’ a uniform interpretation of EEA law and the underlying Community law.
From the point of view of the EFTA Court, it was (and still is) a cause for concern that the ECJ has introduced reciprocity between the contracting parties' obligations as a consideration when interpreting international agreements signed by the EU. There is good reason to believe that the refusal of the EFTA States and their courts to give direct effect to the provisions of their bilateral free-trade agreements with the EC contributed to the ECJ's position in Polydor and Kupferberg. Further, in its judgment in Portugal v Council of 1999, the ECJ refused to allow the WTO rules direct effect in the EU on the grounds that they did not have the same effect in ‘some of the contracting parties, which are among the most important commercial partners of the Community’. Footnote 10 The ECJ pointed out that the WTO agreements are ‘based on reciprocal and mutually advantageous arrangements’ and that lack of reciprocity with respect to the application of the rules in the contracting parties' internal law could lead to ‘disuniform application of the WTO rules’ (paragraph 45). This approach is relevant in the EEA context, as the fourth paragraph of the preamble states that the EEA Agreement was entered into ‘on the basis of equality and reciprocity and of an overall balance of benefits, rights and obligations for the Contracting Parties’. During the EEA negotiations it was important for the EC and its Member States to ensure that the EFTA states could not choose the benefits with full market rights but reject the obligations that the EU Member States have to live with (‘cherry picking’).Footnote 11
This reciprocity consideration links the effect of EEA law in the EU to the sensitive issue of the effect of EEA law in the legal orders of the dualistic EFTA States. The EFTA Court appears to have been far more aware of this link than has academic proponents. Tellingly, it is primarily authors with connections to the EFTA Court—including the Court's former president Sevón, former judge Norberg and current president Baudenbacher—who have introduced reciprocity into the interpretation of EEA law.Footnote 12 For the EFTA Court, the challenge has always been to realise the goal of a homogeneous EEA by balancing the EU and its Member States' legitimate expectations of reciprocity against the (Nordic) EFTA States' precondition that the EEA Agreement should not affect their dualistic approach to the relationship between international obligations and national law. As Sevón warned, if the ECJ were to deem the legal protection of the market operators in the EFTA pillar of the EEA to be unequal to that offered in the EU, there would be a risk of Polydor and Kupferberg gaining new currency.Footnote 13
III. STARTING POINT: A SCEPTICAL ECJ
The task of the EFTA Court was not made easier by the fact that, even before the EEA Agreement entered into force, the ECJ had openly expressed its scepticism about the possibility of realising the goal of uniform interpretation of EEA law and the underlying Community law. In Opinion 1/91 regarding the original draft of the EEA Agreement, the ECJ emphasized the differences between the EEA Agreement's limited objective of market access and the EEC Treaty's objective of a European Union. Further, it stressed the difference between the natures of the two treaties, describing the EEA Agreement as a treaty that ‘essentially only creates rights and obligations between the Contracting Parties, and provides for no transfer of sovereign rights to the inter-governmental institutions which it sets up’, in contrast to the EEC Treaty, which was heralded as ‘the constitutional charter of a Community based on the rule of law’ (paragraph 20). After rejecting the idea that the linkage to earlier ECJ case-law pursuant to Article 6 EEA would suffice to ensure future legal homogeneity, the ECJ concluded categorically that ‘the divergences which exist between the aims and context of the agreement, on the one hand, and the aims and context of Community law, on the other, stand in the way of the achievement of the objective of homogeneity in the interpretation and application of the law in the EEA’ (paragraph 29).
It has rightly been pointed out that the ECJ was clearly opposed to a competing EEA Court,Footnote 14 and that this was presumably why Opinion 1/91 focused rather one-sidedly on the differences between Community law and EEA law.Footnote 15 The following year, when the ECJ in Opinion 1/92 accepted the re-negotiated version of the EEA Agreement, the tone was friendlier.Footnote 16 The EFTA Court must nevertheless have noted that the ECJ pointed out that the ‘divergences’ still remained between the homogeneity objectives on the one hand and the differences in purpose and context between the EEA Agreement and Community law on the other (paragraphs 17–18). The ECJ's conclusion that the amended Agreement was compatible with Community law was not due to a change of mind about the possibility of ensuring uniform interpretation of EEA and Community law, but solely to the fact that there was no longer a common EEA Court that could hamper the ECJ's future interpretation of Community law. The ECJ thus stressed that the EFTA Court ‘will exercise its jurisdiction only within EFTA’ (paragraph 19). This was hardly to be understood as an invitation to constructive judicial collaboration.
The EFTA Court's position was further weakened, if not legally then at least in practical terms, when the Swiss electorate voted against EEA participation in a referendum in 1992 and even further when Finland, Sweden and Austria joined the EU in 1995, just a year after the EEA Agreement finally came into force. As regards the relative strength between the two EEA pillars, the ratio has widened from an assumpted 12:7 when the ECJ published its Opinion 1/92, to 15:3 in 1995, and even further to the present 27:3. Instead of a medium-sized court taking cases from seven EFTA States, the EFTA Court ended up with a minimum of three judges, jurisdiction over one small and two tiny States and a disturbingly small number of cases. Moreover, considering the fact that appointment and reappointment of the judges lies in the hands of the Governments of the EFTA States alone (article 30 SCA), the independence of the EFTA Court might be called into question if it were to be perceived as more sympathetic to the arguments of government lawyers than is the ECJ.
IV. THE STRATEGY OF THE EFTA COURT: DYNAMIC INTERPRETATION OF EEA LAW
A. Ensuring the Effectiveness of EEA Law in the EFTA States
In the EFTA Court's very first case, Restamark, the Commission argued in favour of direct effect of the EEA Agreement.Footnote 17 However, since the main part of the Agreement had already been implemented in Finnish law, it sufficed for the Court to point out that the Agreement's Protocol 35 must be interpreted such that individuals are entitled to invoke and claim at national level any rights that could be derived from implemented provisions of the EEA Agreement, provided they are unconditional and sufficiently precise.Footnote 18 Referring to the ECJ's conclusion that the corresponding provision regarding State monopolies of a commercial character in article 31 EC (now 37 TFEU) satisfied these criteria, the EFTA Court held that this should apply mutatis mutandis to article 16 EEA. In the Restamark case, the consequence of this was that a private importer was allowed, as is the case under EU law, to invoke a provision in the main part of the EEA Agreement before a national court, setting aside national legislation that established a monopoly on the import of alcoholic beverages. Since the main part of the EEA Agreement had also been implemented en bloc in the other dualistic EFTA States, the parity of the contracting parties' obligations had been fully safeguarded regardless of the issue of direct effect. Tellingly, Baudenbacher has stated on a number of occasions that in Restamark the EFTA Court established ‘quasi-direct effect’ for the main part of the EEA Agreement.Footnote 19
In the Einarsson case in 2002, the EFTA Court followed up the Restamark approach by pointing out that Protocol 35 EEA requires the EFTA States, where necessary, to adopt a special statutory provision ensuring the primacy of implemented EEA provisions that confer rights on individuals in a sufficiently precise and unconditional manner over other rules of law.Footnote 20 Protocol 35 is not as such covered by the national implementation of the main part of the EEA Agreement in the dualistic EFTA States, but must be deemed to have been implemented in that the dualistic EFTA States have adopted such primacy rules as prescribed in the Protocol.Footnote 21 With reference to the ECJ's interpretation of the prohibition of discriminatory taxation in article 90 EC (now 110 TFEU), the EFTA Court found that the corresponding provision in article 14 EEA was sufficiently precise and unconditional to entitle a private importer of foreign books to Iceland to demand that it should take precedence over conflicting national rules prescribing different taxes for Icelandic and foreign books. Through the national implementation of Protocol 35, we can, again in the words of Baudenbacher, assert that in Einarsson the EFTA Court established ‘quasi-primacy’ for implemented EEA provisions.Footnote 22
In Restamark and Einarsson, the EFTA Court managed to keep the balance between the obligations of the contracting parties without impinging on the Nordic EFTA States' precondition that the EEA Agreement should not undermine their dualistic approach to the relationship between international obligations and national law. Far more controversial was the conclusion in Sveinbjörnsdóttir that the EEA Agreement embraces a principle of State liability for breach of EEA obligations.Footnote 23 Where the breach consists of an EEA provision not being (correctly) implemented in national law, one cannot disregard the fact that State liability means that EEA rules that are not operative as such in national law nevertheless have internal legal effect.
The EFTA Court found justification for EEA State liability in the homogeneity objective, the objective regarding effective protection by law of rights that the Agreement assigns to individuals and the duty of loyalty pursuant to article 3 EEA.Footnote 24 Opinion is divided as to how convincing the EFTA Court's argumentation is on this point, but there can be no doubt that it constitutes a clear example of judicial activism.Footnote 25 Its relevance in the present context is that extending the existence of State liability to the EFTA States contributes to a greater degree of reciprocity between the obligations of the Contracting Parties. From the point of view of the EU Member States, the risk of liability as a result of breach of EU law obligations is a price they have to pay for membership of the EU. Since the EEA Agreement forms an integral part of EU law, the Member States—regardless of the existence of a specific EEA liability—may also, within the scope of the EEA Agreement, incur liability vis-à-vis market operators or individuals residing in one of the EFTA States.
Thus, consideration of the overall balance in the obligations of the contracting parties seems to be a key factor in understanding the EFTA Court's effect-oriented approach to the homogeneity objective in Sveinbjörnsdóttir. The introduction of the EEA State liability shows clearly that in the opinion of the EFTA Court the homogeneity objective is not limited to the interpretation of the substantive content of the EEA provisions, but also embraces their legal effects in national law.Footnote 26 Moreover, Sveinbjörnsdóttir shows that the EFTA Court also strives for homogeneity of effect in cases in which the argumentation leading up to the corresponding result in EU law is not fully applicable in the EEA context. In Sveinbjörnsdóttir, a direct transfer of the EU law principle of State liability appeared problematic, as it had been closely linked by the ECJ to the (then) European Community a supranational legal order.Footnote 27 The EFTA Court circumvented this link by establishing a specific EEA liability,Footnote 28 simultaneosly safeguarding the Agreement's homogeneity objective by ensuring that the content of the liability corresponds to that of the Member States in the EU.
Neither in Sveinbjörnsdóttir nor in subsequent judgments has the EFTA Court given any real justification for its effect-oriented understanding of the homogeneity objective. The Court's President has, however, on several subsequent occasions explicitly linked the principle of State liability to the overall balance between the obligations of the Contracting Parties.Footnote 29
A further step in the ‘supranational’ direction was taken in the Karlsson judgment in 2002, where the EFTA Court stated that EEA law imposes a duty of EEA-consistent interpretation of national law.Footnote 30 This standpoint means that the EEA Agreement as such Footnote 31 obliges national courts to also take account of EEA rules that are not operative in national law. The duty of EEA-consistent interpretation is limited by ‘the interpretative methods recognised by national law’,Footnote 32 but this applies mutatis mutandis to the duty of consistent interpretation of national law in the EU.Footnote 33 The point in the present context is that, here too, reciprocity is established between the obligations of the Contracting Parties.
The EFTA Court's dynamic development of the EEA Agreement found its limits in the matter of direct effect and primacy of incorrectly implemented directives. It is, however, perhaps the Court's handling of this question that evidences most clearly its awareness of how interpretation of the EEA Agreement may be perceived by the ECJ. Tellingly, several of the authors who argued that the principles of direct effect and primacy principles ought to be seen as an integral part of the Agreement linked their view to the requirement of an overall balance between the objectives of the Contracting Parties.Footnote 34 The issue first came to a head in the case Criminal proceedings against A in 2007, but the EFTA Court had already five years earlier in an obiter dictum in Karlsson expressed clearly that it followed from the Agreement's article 7 and Protocol 35 that the principle of direct effect was not part of the EEA Agreement.Footnote 35 In Criminal proceedings against A, the EFTA Court could simply have made a brief reference to the opinion in Karlsson, and added that EEA law consequently does not require non-implemented EEA rules to take precedence over national rules. It is therefore remarkable that the EFTA Court found it appropriate to emphasise not only the homogeneity objective, the EFTA States' implementation obligations under article 7 EEA and the duty to give primacy to implemented EEA rules pursuant to Protocol 35, but also the duty of EEA-consistent interpretation of national law, the EFTA States' liability for violations of EEA law and the power of the EFTA Surveillance Authority (ESA) under article 31 SCA to bring a case concerning a violation of EEA law before the EFTA Court.Footnote 36 One is left with the impression that the EFTA Court's reasoning was addressed just as much to the EU (ie the ECJ) as to the Liechtenstein court that had submitted the case. With the exception of the duty of EEA-consistent interpretation of national law, the latter was not the appropriate addressee for the obligations and remedies listed. The message from the EFTA Court seems to be that the EU need not worry about the absence of direct effect and primacy in the EFTA pillar since EEA law contains a number of other mechanisms intended to guarantee a homogeneous EEA and an overall balance between the obligations of the Contracting Parties.
B. Minimizing the Importance of Differences in Context and Purpose
In addition to ensuring that EEA law becomes effective in the EFTA States' national laws, the EFTA Court must have been anxious to prove wrong the ECJ's forecast that differences in context and purpose would inevitably undermine the objective of an interpretation of EEA rules in conformity with identically worded provisions of EU law. Were it to become widely perceived in the EU that the EFTA Court—with reference to the EEA Agreement's more limited scope and less comprehensive objectives—was more sympathetic towards attempts by the EFTA States to defend their political latitude than would be the case in the ECJ, the condition of reciprocity between the Contracting Parties' obligations could be expected to come under fire. At worst, the EFTA Court could be perceived by the ECJ as a contributor to the feared ‘cherry-picking’ of the EFTA States rather than as an independent court of justice.
As early as in the Restamark case, the Finnish government argued that the absence of harmonisation of taxation rules in the EEA could justify the existence of a State import monopoly for alcoholic beverages, even though such a monopoly might be incompatible with Community law.Footnote 37 This attempt to legitimize the import monopoly as a system for levying taxes seemed rather contrived. The argument was not even mentioned by the EFTA Court, which asserted instead that the EEA Agreement's provisions concerning quantitative restrictions on imports and commercial monopolies in articles 11, 13 and 16 must be interpreted in accordance with the identically worded provisions in articles 28, 30 and 31 EC (now 34, 36 and 37 TFEU).Footnote 38
What is more surprising is that the EFTA Court itself, in the Maglite case from 1997, introduced into its interpretation of article 7(1) of the Trademark Directive (89/104/EEC) the argument that the EEA Agreement—unlike the then EC Treaty—does not establish a customs union with a common commercial policy in relation to third countries.Footnote 39 In the Silhouette case the following year, it became evident that the ECJ did not share the EFTA Court's understanding that the Directive allows for international exhaustion of trademark rights.Footnote 40 The EFTA Court made a U-turn when the issue was raised again in the L'Oréal case of 2008. The Court pointed out that the homogeneity principle entails a presumption that provisions worded in the same manner in the EEA Agreement as in Community law shall be interpreted in the same manner, even though differences in scope and purpose may in ‘specific circumstances’ lead to differences in interpretation.Footnote 41 The governments of all of the three EFTA States as well as the ESA argued that the absence of a common commercial policy justified the Maglite solution, but the EFTA Court concluded that the differences between the EEA Agreement and the (then) EC Treaty with respect to commercial relationships with third countries did not constitute ‘compelling reasons’ for divergent interpretations of the Trademark Directive in EEA law and Community law.Footnote 42 Read in context, L'Oréal leaves one with the impression that the EFTA Court not only departed from the concrete solution in Maglite, but also raised the threshold for allowing differences in purpose and context to justify derogation from the homogeneity principle.Footnote 43 It must be said, however, that the EFTA Court was in a difficult position; in Silhouette the ECJ had not only interpreted the Trademark Directive differently from the EFTA Court, but also assumed that its interpretation was valid in the context of EEA law.Footnote 44 In a case where the ECJ had disregarded the specific EEA arguments put forward by the EFTA Court in Maglite, it would not look too good if the EFTA Court were to adhere to a divergent interpretation that gave the EFTA States greater political latitude than the EU States.Footnote 45
The minimising by the EFTA Court of the importance of differences in context and purpose had, however, started well before L'Oréal. In the Rainford-Towning case from 1998, Liechtenstein's government had argued—referring to Maglite and ECJ Opinion 1/91—that ‘the lesser ambitions of the EEA’ indicated that the elimination of restrictions on the freedom of establishment in relation to article 31 must be perceived as less absolute than the ECJ's interpretation of the corresponding provision in article 43 EC (now 49 TFEU).Footnote 46 The EFTA Court was not convinced.
Correspondingly, in case E-6/98 Norway v ESA, the Norwegian government argued in vain that a system of differentiated social security contributions paid by employers did not constitute State aid since the EEA Agreement does not aim at harmonisation of taxation.Footnote 47 As the ECJ had previously held that the tax-law character of an aid system is no obstacle to regarding it as State aid in relation to article 87 EC (now 107 TFEU), the Norwegian government in fact argued that the identically worded provision in article 61 EEA should be interpreted differently as a consequence of the differences in the context of the provisions. The absence of taxation harmonisation as a justification for interpreting EEA law provisions differently from identically worded provisions in Community law was also argued by the Icelandic government in Einarsson Footnote 48 and again by the Norwegian government in Fokus Bank Footnote 49—the EFTA Court was not convinced in either case. The fact that tax law as such is not covered by the EEA Agreement does not prevent the Agreement's basic rules on State aid and free movement of goods, persons, services and capital placing the same restrictions on the EFTA States' taxation competence as identically worded provisions in EU law.
Another difference between EEA law and EU law relates to the EFTA States' insistence that the Agreement should not impinge on the independence of the national courts. In the Piazza case in 2005, the Liechtenstein government argued that in consideration of the independence of the national courts, EEA law could not impact on the EFTA States' procedural legislation in the same way as Community law.Footnote 50 However, the EFTA Court made no mention of this argument when it established that article 40 EEA concerning the free movement of capital must be interpreted in conformity with article 56 EC (now 63 TFEU), with the consequence that EEA law prohibits procedural legislation that accepts the provision of security for procedural costs from national sources only.Footnote 51
Thus, it was no surprise when the EFTA Court in case E-2/06 ESA v Norway, concerning the conditions for concession for acquisition of hydropower resources, rejected the arguments of the Norwegian and Icelandic governments claiming that article 125 EEA ought to be interpreted more narrowly than the identically worded provision in article 295 EC (now 345 TFEU) ‘due to fundamental differences between the EC Treaty and the EEA Agreement’.Footnote 52 The Norwegian government supported this argument with reference to the ECJ Opinion 1/91.Footnote 53 The EFTA Court reiterated the wording from the above-mentioned Rainford-Towning case that only in ‘specific circumstances’ may differences in scope and purpose lead to divergent interpretation. In the Court's opinion there were no such circumstances in this case. Article 125 EEA was therefore to be interpreted in conformity with article 295 EC (now 345 TFEU), with the consequence that national rules on property ownership are not excluded per se from the scope of the EEA Agreement.Footnote 54
Since the ‘compelling reasons’ test in the L'Oréal case, the threshold for divergent interpretations seems to have been set even higher. Thus, we can safely assume that it will be very difficult for the EFTA States to advance differences in context and purpose between EEA law and EU law in order to secure greater political leeway than the ECJ allows the EU Member States.
C. Rejection of the EFTA States' Preconditions when Signing the EEA Agreement
In addition to arguments relating to differences in purpose and context, the EFTA States have on several occasions referred to their preconditions when signing the EEA Agreement (or in subsequent decisions in the EEA Joint Committee) to argue in favour of interpreting EEA provisions differently from the ECJ's interpretation of identically worded provisions of EU law. In view of the limits of the homogeneity objective defined in L'Oreál, the question is whether the preconditions of the contracting parties in a concrete case constitute ‘compelling reasons’ for an interpretation of EEA law that derogates from EU law. The EFTA Court has never accepted such objections. When analysing the case-law, there is nevertheless reason to distinguish between unilateral preconditions and preconditions claimed to be common to all Contracting Parties. Further, a distinction also has to be drawn between preconditions relating to the interpretation of substantive EEA law and those relating to the effect of EEA law in national law.
It is obvious that no significant importance can be attached to preconditions that cannot even be claimed to have been common to all contracting parties at the date of signature. Already in the Restamark case, the Finnish government tried unsuccessfully to save its import monopoly for alcoholic beverages by referring to a joint declaration that the Nordic EFTA States had tied to the EEA Agreement, justifying Nordic monopoly systems on important grounds of health and social policy.Footnote 55 The EFTA Court judgment in the aforementioned case on acquisition of hydropower resources from 2007 is also illustrative of this point. This judgment dismissed objections from the Icelandic government that for both Norway and Iceland it was ‘a pre-condition for the Agreement […] that ownership over energy resources would remain unaffected’, stating that ‘[u]nilateral expressions of understanding of the kind claimed to have been made by Norway and Iceland’ could not justify an interpretation of Article 125 EEA that diverged from the ECJ's interpretation of article 295 EC (now 345 TFEU).Footnote 56
The evaluation is more difficult in cases where common preconditions that would lead to a divergent interpretation of EEA law are claimed to be present. If we—as the ECJ did in Opinion 1/91—base ourselves on general rules of international law relating to the interpretation of treaties, it is undisputed that treaty provisions should be interpreted in accordance with the common intentions of the Contracting Parties. When interpreting the substantive content of the EEA rules, however, acceptance of such arguments would conflict with the overriding objective of a homogeneous EEA per se.
The judgment in the ESA v Liechtenstein case of 2007 is illustrative of the EFTA Court's attitude to this issue. The Liechtenstein government argued in this case that it was ‘a condition sine qua non when acceding to the EEA’ that a specific form of allowance for persons in need of care was not covered by Regulation 1408/71 relating to the application of social security schemes to employed persons, to self-employed persons and to members of their families moving within the EU, and that it could not therefore be claimed by EEA citizens who are not resident in Liechtenstein.Footnote 57 In connection with Liechtenstein's accession to the EEA Agreement, the allowance in question was listed in a special Annex to the Regulation, which was a precondition for excluding it. However, subsequent ECJ case-law made it clear that such listing is not in itself sufficient to exclude a benefit if it did not also meet the Regulation's own conditions for such exclusion (which the Liechtenstein allowance failed to do).Footnote 58 Liechtenstein objected that the inclusion of the allowance in the Annex was the result of an EEA Joint Committee decision and therefore subject to ‘the rules of public international law’, implying that it was ‘binding upon the parties to it and must be performed by them in good faith’.Footnote 59 This international law approach, however, gained no backing from the EFTA Court, however, which stated that ‘in its interpretation of the EEA Agreement, the Court cannot be bound by mere expectations of the Contracting Parties as to the exact content of the obligations the Parties enter into.’Footnote 60
Even if we accept that specific preconditions of the Contracting Parties must give way to the overriding intention of a homogeneous EEA when interpreting the substantive content of the EEA rules, it is not given that this applies mutatis mutandis to the legal effect of the EEA rules in national law. In Sveinbjörnsdóttir, Iceland, Norway, Sweden and—interestingly—the Commission all rejected the existence of State liability, referring explicitly to the common intentions when negotiating the Agreement.Footnote 61 The EFTA Court, however, made no mention of this objection. The Norwegian government reiterated the argument in Karlsson,Footnote 62 but once more the EFTA Court failed to comment. Of course, the outcome of the cases reveals that the Court did not attach much importance to the intentions of the Contracting Parties, but a justification of this would definitely have been appropriate. In the absence of such justification, we can but note that the EFTA Court seems to share its President's opinion that the initial intentions of the Contracting Parties to the EEA Agreement has ‘no relevance’ to the development of EEA law, at least not to the interpretation of the main part of the Agreement.Footnote 63
D. Rejection of all Attempts by the EFTA States to Increase their Political Latitude
It seems that the EFTA Court has deemed it important not to appear more sympathetic than the ECJ towards attempts by the EFTA States to increase their political latitude also in cases in which arguments for strict interpretation of EEA law are not linked to EEA-specific circumstances and can thus be claimed to apply to interpretation of the underlying EU law as well.
An example of this can be seen in the ‘Postdoc’ case from 2003, in which the Norwegian government requested the EFTA Court to derogate from the ECJ's interpretation of the Employment Equality Directive (76/207/EEC), so that affirmative action would no longer be regarded as discrimination in the sense of the Directive, but rather as ‘an intrinsic dimension of the very prohibition thereof’.Footnote 64 This would have made it possible to earmark certain university posts for women, thereby increasing the political latitude of the EFTA States. However, the EFTA Court dismissed the request, referring to the homogeneity principle.Footnote 65
The EFTA Court has been similarly dismissive of attempts to extend the range of legitimate considerations that can be used to justify restrictions to the four freedoms.Footnote 66 For example, in Einarsson the Icelandic and Norwegian governments argued unsuccessfully that the preservation of the Icelandic language as a central component of Iceland's cultural heritage and national identity should be allowed to justify derogation from the prohibition against discriminatory taxation in article 14 EEA.Footnote 67
Two cases deserving of particular attention in this connection are Nille from 1997 and Fokus Bank from 2004. In Nille, the Norwegian government argued that the possible discriminatory effect of Norwegian provisions on the rental of video cassettes could be justified by reference to the rules' cultural policy objectives.Footnote 68 The EFTA Court dismissed this on the grounds that cultural policy is not listed among the legitimate considerations in article 13 EEA, and that the ECJ had consistently held that the corresponding provision in article 36 EC (now 42 TFEU) must be interpreted strictly, as it constitutes a derogation from the basic rule that all obstacles to the free movement of goods between the Member States must be eliminated.Footnote 69 Shortly afterwards, however, the ECJ made it clear in de Agostini that the right to supplement the then EC Treaty with ‘overriding requirements of general public importance'is precluded only in cases of direct discrimination.Footnote 70
A similar discrepancy between EFTA Court and ECJ case-law came to light in Fokus Bank. The Norwegian government requested that the EFTA Court accept a justification of the restriction on the free movement of capital in the EEA on the grounds of cohesion of the international tax system, which would mean that the effect of international tax treaties must be included when determining the presence or absence of discrimination between resident and non-resident shareholders.Footnote 71 The EFTA Court rejected this on the grounds that an EEA Member State cannot make the rights pursuant to article 40 EEA contingent on the content of a bilateral agreement entered into with another EEA Member State—such a solution would give bilateral tax agreements precedence over EEA law.Footnote 72 In the subsequent ACT Group Litigation and Denkavit cases, however, the ECJ arrived at the opposite conclusion.Footnote 73 In his opinion in ACT Group Litigation, AG Geelhoed took a clear stand against the EFTA Court's solution in Fokus Bank, and pointed out that ‘[i]f the effect of the DTC [double taxation convention] in an individual case were not taken into account, this would ignore the economic reality of that taxable subject's activity and incentives in a cross-border context’.Footnote 74
With hindsight, it is easy to say that the EFTA Court in Nille and Fokus Bank failed to predict subsequent development in ECJ case-law. However, in light of the homogeneity objective and the requirement for reciprocity between the Contracting Parties' obligations it is hardly surprising that the EFTA Court is reluctant to forestall a development in ECJ case-law that would give the Member States greater political leeway. Nonetheless, this begs the question whether the EFTA Court is as cautious when the opposite is the case, ie when it is a matter of developing EEA law in a more integration-friendly direction.
E. Integration-friendly Interpretation in ‘Hard Cases’
In cases in which there are no EEA-specific circumstances to justify a divergent interpretation, it is fairly clear that the homogeneity objective presupposes an interpretation of EEA law in conformity with the underlying EU law. However, where the interpretation of EU law is unclear, the objective of homogeneous interpretation as such provides little guidance.
In particular, the Kellogg's case from 2001 shows that the EFTA Court does not shy away from developing EEA law in an integration-friendly direction. At issue in this case was whether a prohibition on the import of cornflakes fortified with vitamins and minerals was in accordance with the provisions of article 11 EEA. The Norwegian government pointed out that the ECJ in Sandoz had stated that the Member States were only obliged to allow imports of foodstuffs fortified with vitamins if the additives filled a ‘real nutritional need’.Footnote 75 In Kellogg's, however, the EFTA Court did not accept that the mere absence of a nutritional need in the population was sufficient to justify an import ban.Footnote 76 On the contrary, the authorities were required to conduct a specific assessment of whether the vitamin and iron-enriched cornflakes would present any danger to public health. In this case, the EFTA Court's approach entailed a clear restriction of the EEA States' right to justify import restrictions by reference to the precautionary principle, inter alia in the form of a requirement for a comprehensive assessment of the risk to public health based on the most recent scientific information.Footnote 77 Shortly afterwards, in a similar case against Denmark, the Commission referred to the EFTA Court judgment and described it as an expression of a legal development that meant that the ECJ's previous approach in the Sandoz judgment should be deemed to have been abandoned. The Danish government replied that the EFTA Court's judgment was inconsistent with existing ECJ case-law. Referring explicitly to Kellogg's, the ECJ decided to follow the EFTA Court.Footnote 78
Further, two interesting examples in private-law disputes are provided by the LO case of 2002 and the Paranova case of 2003. A central issue in LO was whether collective agreements between employers' and employees' associations are included in the prohibition of cartels in article 53 EEA. Referring to the ECJ's Albany judgment from 1999,Footnote 79 the Icelandic, Norwegian and Swedish governments argued that collective agreement provisions aimed at improving working and employment conditions for the employees must, by virtue of their nature and purpose, fall outside the scope of article 81 EC (now 101 TFEU), and this should apply mutatis mutandis for article 53 EEA.Footnote 80 The Commission and the ESA also argued in favour of a limited judicial review of the collective agreement. The EFTA Court, however, pointed out that the immunity of collective agreements from rules of competition cannot be given unlimited application.Footnote 81 In this connection, the Court referred to AG Jacobs' opinion in Albany, in which he argued that collective agreements cannot deprive the prohibition on cartels of all its meaning.Footnote 82 The reference is interesting because the ECJ did not refer to the Advocate General in this case. The EFTA Court subsequently arrived at the conclusion that even though it is to be presumed that disputed provisions in the collective agreement fell outside the scope of article 53 EEA, the national court would nevertheless have to investigate how the agreement was being practised, whether the parties actually pursued the stated socio-political objectives, and also what effects the agreement had.Footnote 83 According to Baudenbacher, one of the judges in the case, the Court thus demonstrated a more competition-friendly attitude than the ECJ.Footnote 84
The EFTA Court's competition-friendly attitude also manifested itself in Paranova, a case which dealt with the issue of trademark rights in the parallel import of pharmaceutical products. In Bristol-Myers Squibb, the ECJ had found that a parallel importer may only repackage and reaffix the producer's trademark where it is necessary in order to market the product in the Member State of import.Footnote 85 The issue in Paranova was whether the necessity test should also be applied to the parallel importer's packaging design in a case of lawful repackaging. The EFTA Court rejected this with reference to the fundamental principle of free movement of goods. After having lawfully repackaged the products, and reaffixed the proprietor's trademark, the parallel importer is to be considered an operator on essentially equal footing with the manufacturer and the proprietor.Footnote 86 According to article 7(2) of the Trademark Directive, a ‘legitimate reason’ to oppose further commercialisation of repackaged pharmaceutical products only exists if the packaging design is liable to harm the trademark or the proprietor's reputation. According to Baudenbacher, one of the judges in the case, this judgment illustrated a fundamentally sympathetic attitude towards the free movement of goods in the EFTA Court.Footnote 87Paranova led to the English Court of Appeal in Boehringer Ingelheim II asking the ECJ whether the EFTA Court's understanding of the Directive was correct.Footnote 88 It was argued before the ECJ that the EFTA Court had attached insufficient importance to the trademark proprietor's right to present the trademark in the manner he wishes.Footnote 89 The ECJ, however, concurred with the understanding of the EFTA Court.Footnote 90
Further examples of the EFTA Court's integration-friendly approach to EEA law are provided by cases such as Mattel/Lego, Finanger, Norwegian Bankers' Association, E-1/05 ESA v Norway (‘life-insurance’) and Ladbrokes.Footnote 91 The EFTA Court's dynamic interpretation in Finanger won subsequent approval from the ECJ in Candolin,Footnote 92 whereas the offensives in Mattel/Lego, Norwegian Bankers' Association and Ladbrokes failed to convince the ECJ in the subesquent cases de Agostini, Ferring and Liga Portuguesa.Footnote 93
To sum up, the overall impression left by the EFTA Court case-law of the past 15 years is that in ‘hard cases’, the Court does not shy away from taking a leading role in the dynamic development of EEA law. As has been shown, the EFTA Court has in several cases won subsequent approbation of the ECJ, while in other cases it has had to (or will have to) correct its own course after the ECJ delivered its opinion. In the present context, the point is that the EFTA Court's decisions follow a fairly clear-cut pattern. Arguably, Maglite is the only ‘hard case’ in which the EFTA Court preferred a solution giving the EFTA States greater political leeway. It must nevertheless be added that, from the global perspective, an opening for national rules on international exhaustion of trademark rights was the most free-trade-friendly and competition-friendly solution.Footnote 94 Besides, as has been shown, Maglite was later overruled by L'Oréal.
F. In Search of a Good Relationship with the ECJ
A final aspect of EFTA Court case-law of interest in the present context is the clear disinclination to engage in any open confrontation with the ECJ. It is illustrative of this that the EFTA Court has never quarrelled openly with the ECJ's characterization of the EEA Agreement in Opinion 1/91, despite the fact that it has been evident ever since Restamark that the EFTA Court does not share the ECJ's perception that the EEA Agreement's purpose and context constitute an obstacle to the realization of a homogeneous EEA. The Court's silence on this matter is particularly evident in Sveinbjörnsdóttir, where the Icelandic, Norwegian and Swedish governments had all dismissed the existence of State liability for breach of the EEA Agreement, with explicit reference to the ECJ's characterization of the Agreement as an international law agreement that ‘only creates rights and obligations between the Contracting Parties’, unlike Community law ‘the subjects of which comprise not only Member States but also their nationals’.Footnote 95 The EFTA Court's attempt to avoid open confrontation with the ECJ on this point is probably not surprising; it is nevertheless clear that the recognition of private individuals as legal persons under the EEA Agreement constitutes a sine qua non condition for the existence of State liability.Footnote 96 Rather than ignoring the ECJ's opinion, the EFTA Court ought to have pointed out in Sveinbjörnsdóttir that Opinion 1/91 was given on the basis of an earlier draft of the EEA Agreement that did not include the relatively unambiguous wording of the eighth paragraph of the preamble on ‘the important role that individuals will play in the European Economic Area through the exercise of the rights conferred on them by the Agreement and through the judicial defence of these rights’.Footnote 97 In this way, the Court could have managed quite easily to steal a march on its critics.
This striving of the EFTA Court towards a good relationship with the ECJ is also reflected in what can be described as a fairly active quest for support in ECJ case-law. A good example is the use of the ECJ's reference to Sveinbjörnsdóttir in the Rechberger case. When the issue of EEA liability re-emerged in the Karlsson case, the EFTA Court took the opportunity to point to this reference in a manner that indicated that the ECJ agreed with the EFTA Court's understanding of State liability as an integral part of the EEA Agreement.Footnote 98 As we will see below, opinion is divided on whether the ECJ's reference to Sveinbjörnsdóttir can really be interpreted in this way. For our present purposes, the point is that in Karlsson the EFTA Court obviously wanted to take the ECJ ‘onboard as an ally’.Footnote 99
V. THE TRANSFORMATION: A FRIENDLIER ECJ
In the years immediately following the entry into force of the EEA Agreement, the ECJ's attitude seemed rooted in the scepticism it had manifested in Opinion 1/91. An early example of this can be seen in AG Fennelly's opinion from 1996 in Yamanouchi Pharmaceutical, in which the plaintiff argued that certain specific changes in the EEA version of a Regulation on protection certificates for medicinal products shed light on the interpretation of the original EC version of the Regulation. This argumentation was somewhat contrived—it was evident that the sole purpose of the changes in the EEA version of the Regulation was to assimilate the EFTA States into the existing certificate system. However, the Advocate General found it appropriate not only to dismiss the argument, but also to add—with reference to ECJ Opinion 1/91—that an interpretation of Community law in the light of EEA law was always problematic because ‘the differences in character between the EEC/EC Treaty and the EEA Agreement are notorious.’Footnote 100
Also with respect to its relationship with the EFTA Court, the attitude of the ECJ in the years immediately after the entry into force of the EEA Agreement seemed marked by the reluctance manifested in Opinion 1/91 to allow other courts to influence the development of Community law. It is illustrative that the ECJ in Franzén deliberated the Swedish import and retail monopoly for alcoholic beverages with absolutely no mention of the EFTA Court's judgment in Restamark.Footnote 101 Similarly in Silhouette, the ECJ elected to tacitly ignore the EFTA Court's interpretation of the Trademark Directive in Maglite, despite the fact that AG Jacobs had referred to the case in his opinion.Footnote 102 Thus, Süzen and de Agostini, both from 1997, represented an important break in that they were the first cases in which the ECJ, through its references to Ulstein and Røiseng and to Lego/Mattel respectively, referred to EFTA Court case-law.Footnote 103
The matter of the EEA Agreement's position in Community law came to a head before the Court of First Instance (CFI) in the Opel Austria case in 1997.Footnote 104 With the support of the Austrian government, Opel Austria brought an action against the Council for the annulment of a regulation that introduced a customs duty on a specific type of gearbox manufactured by a factory in Austria. A central question in this case was whether the regulation was an infringement of article 10 EEA. With reference to ECJ Opinion 1/91, the Council, with the support of the Commission, argued that the differences between the EEA Agreement and the then EC Treaty indicated that article 10 EEA could not be interpreted in conformity with the identically worded prohibition on customs duties in the EC Treaty.Footnote 105 However, in a decision whose importance to EEA law cannot be overestimated, the CFI stated that from the date of its entry into force the EEA Agreement constituted an integral part of Community law and that the prohibition in article 10 was so unconditional and precise that it had direct effect.Footnote 106 With reference to article 6 EEA, article 10 was interpreted in conformity with ECJ case-law prior to the date of signature. Further, the CFI held that the EEA Agreement ‘involves a high degree of integration, with objectives that exceed those of a mere free-trade agreement’ and explicitly rejected that ECJ Opinion 1/91 was an argument against uniform interpretation of the EEA Agreement and Community law.Footnote 107 In recognition of the EFTA Court, its judgments in Restamark and Scottish Salmon Growers Association were used to argue that the obligation to consider subsequent ECJ case-law in relation to Article 3(2) SCA would help to attain the homogeneity objective also in the future.Footnote 108 The Council appeared convinced by the argumentation—at least to the extent that it did not appeal the judgment to the ECJ.
The absence of an appeal in Opel Austria meant that the ECJ's position on the EEA Agreement had still not been clarified by the time of the Andersson and the Rechberger cases, in which the Court was asked, respectively, about Swedish and Austrian State liability for breaches of the EEA Agreement in the period in which they were still members of EFTA. In both instances the ECJ declined to answer, stating that it had no jurisdiction to pronounce on the effect of the EEA Agreement in Sweden and Austria prior to their accession to the EU.Footnote 109 In Andersson, however, AG Cosmas took up the alternative question of whether the EEA Agreement embraces a principle of State liability for breach of the Agreement. His opinion was submitted on 19 January 1999, just over a month after the EFTA Court had declared the existence of such a principle in Sveinbjörnsdóttir. However, the Advocate General did not mention the EFTA Court's judgment when, with reference to ECJ Opinion 1/91, he rejected the argument that the Community law principle of liability for Member States for breach of Community law could be regarded part of EEA law.Footnote 110 According to the Advocate General, it followed implicitly from the ECJ Opinion that the principles of primacy and direct effect of Community law did not apply to EEA law, which meant that State liability, which he described as ‘inextricably linked to the fundamental principles set out above’, could not be transposed to the EEA either.
When judgment was handed down in Andersson, the ECJ did not mention the Advocate General's views on the liability issue. However, in its judgment in Rechberger of the very same day, the ECJ referred in an obiter dictum to the EFTA Court's position in Sveinbjörnsdottir.Footnote 111 The ECJ did not explicitly state its own view on the liability issue. It would, however, be rather odd if the ECJ were to have made the Austrian court explicitly aware of the EFTA Court's position on the issue if it disagreed with it. The absence of a corresponding reference in Andersson makes it obvious that the inclusion of the reference in Rechberger had been carefully thought through. Both cases were decided in plenary session, but with somewhat differing compositions.Footnote 112 However, seven of the judges participated in both judgments and the Rechberger judgment refers explicitly to the decision on the jurisdiction issue in Andersson. The judges must, therefore, have been well aware of the difference between the judgments regarding the reference to the EFTA Court—a difference that can only be explained by the different composition of the Court in the two cases. The predominant opinion in the literature is therefore that the reference in Rechberger has to be interpreted as an implicit recognition of the EFTA Court's position.Footnote 113 Viewed in conjunction with the silence in Andersson (and AG Cosmas' dismissive attitude), it seems evident that at this point there was some internal disagreement in the ECJ regarding the interpretation of the EEA Agreement.
In the wake of Rechberger there were several judgments in which the ECJ referred to EFTA Court case-law to support its arguments concerning interpretation of Community law. In the Oy Liikenne case from 2001, the ECJ concurred with the EFTA Court in Eidesund and Ask that the fact that an undertaking is transferred as a consequence of a public tender procedure does not in itself exclude the transfer from the scope of the Transfer of Undertakings Directive (77/187/EEC).Footnote 114 In Monsanto Agricoltura and the aforementioned case C-192/01, Commission v Denmark, both from 2003, the ECJ agreed with the EFTA Court's interpretation of the precautionary principle in Kellogg's.Footnote 115
There was, however, no final clarification of the position of the EEA Agreement in Community law until the Ospelt case in 2003. A key question in this case was whether Austria could justify restrictions on the free movement of capital in the EEA on the grounds that the EFTA State Liechtenstein was a ‘third country’ under article 57 EC (now 64 TFEU). The ECJ pointed out, however, that the legal basis for considering the movement of capital between an EU Member State and an EFTA State is article 40 EEA, and Annex XII to the Agreement. With reference to the fact that ‘[o]ne of the principal aims of the EEA Agreement is to provide for the fullest possible realisation of the free movement of goods, persons, service and capital within the European Economic Area, so that the internal market established within the European Union is extended to the EFTA States’, the ECJ stated that ‘(i)t is for the Court, in that context, to ensure that the rules of the EEA Agreement which are identical in substance to those of the Treaty are interpreted uniformly within the Member States.’Footnote 116 The ECJ made no mention at all of Opinion 1/91, but referred instead to the discussion of the EEA Agreement's homogeneity objective in the more EEA-friendly Opinion 1/92. The Austrian government argued unsuccessfully that article 40 EEA should be interpreted in conformity with former article 67(1) EC, with the consequence that subsequent ECJ case-law relating to article 56 EC (now 64 TFEU) was without relevance.Footnote 117 Rejecting this objection, AG Geelhoed stated inter alia that the EFTA Court in Islandsbanki had implicitly based its findings on an interpretation of article 40 EEA in conformity with Article 56 EC.Footnote 118 The ECJ, on the other hand, made no issue at all of the change in the EC Treaty's rules on capital following the Treaty of Maastricht. It simply stated that article 40 EEA must be interpreted in conformity with the ‘largely identical’ provisions of article 56 EC.Footnote 119
The clarification in Ospelt was followed up in the following year in Bellio F.lli. The EFTA Court had already availed itself of the opportunity to refer to Ospelt in the E-1/03 ESA v Iceland and Ásgeirsson cases. The ECJ followed up by stating that ‘both the Court and the EFTA Court have recognized the need to ensure that the rules of the EEA Agreement which are identical in substance to those of the Treaty are interpreted uniformly.’Footnote 120 The ECJ added that article 13 EEA (on the right to place restrictions on the free movement of goods) is identical in substance to article 30 EC (now 36 TFEU).
In the subsequent judgments in Keller Holding and C-345/05 Commission v Portugal from 2006 and C-104/06 Commission v Sweden from 2007, the ECJ assumed without further discussion that the provisions of the EEA Agreement on the free movement of capital and freedom of establishment in articles 28 and 31 must also be interpreted in conformity with articles 39 and 43 EC (now 45 and 49 TFEU).Footnote 121 A corresponding statement regarding the free movement of services (article 36 EEA/article 49 EC (now 56 TFEU)) was made in the judgment in C-522/04 Commission v Belgium from 2007.Footnote 122 In the same year, the ECJ commented briefly in Ludwigs-Apotheke that the rules in articles 11 and 13 EEA on the free movement of goods are ‘essentially identical’ to those laid down by articles 28 and 30 EC (now 34 and 36 TFEU), with the consequence that the conflict found between Community law and a provision in the German law on medicinal products could be transferred directly to EEA law.Footnote 123
All in all, post-Ospelt case-law leaves us with the clear impression that the ECJ's attitude to the EEA structure has changed from considerable scepticism to recognition that it is actually possible to realise the objective of the participation of the EFTA States in the internal market.Footnote 124
In parallel with this change in attitude, references to EFTA Court case-law have gradually evolved into what can only be described as a dialogue between the two EEA courts. By way of example, in 2004 a further reference was made to Kellogg's in case C-41/02, Commission v the Netherlands;Footnote 125 and the Ahokainen case from 2006 included the reference to Restamark that had been absent nine years earlier in Franzén.Footnote 126 In the Grand Chamber judgment in Fidium Finanz in the same year, the ECJ referred to the EFTA Court judgment in Íslandsbanki to support its understanding that in cases relating to both the freedom to provide services and the free movement of capital, a restriction should be considered against the fundamental freedom that is mainly affected.Footnote 127 In addition to this, we also have the aforementioned judgment Boehringer Ingelheim II from 2007, in which the ECJ explicitly concurred with the EFTA Court's interpretation of the Trademark Directive in Paranova.Footnote 128
This impression of a dialogue between the ECJ and the EFTA Court is reinforced if we also include the Advocates General in our assessment.Footnote 129 Since the brief reference by AG Lenz to Lego/Mattel in case C-222/94, Commission v United Kingdom, the Advocates General have so far referred to EFTA Court case-law in 33 different cases.Footnote 130 Here too the dialogue seems to be growing over time—26 out of a total of 33 references are from cases since the turn of the millennium.
Regarding the ECJ's willingness to engage in dialogue with other international courts, it may be claimed that its volte-face after Opinion 1/91 already happened in P v S from 1996, in which the ECJ made its first reference to the case-law of the European Court of Human Rights (ECtHR).Footnote 131 Once the ECJ had relinquished its opposition in principle to allowing itself to be influenced by the case-law of other courts, it would appear somewhat contrived if it were not also to take account of EFTA Court case-law.Footnote 132 Although the EFTA Court was probably helped along by the growing dialogue between the ECJ and the ECtHR, the ECJ's frequent references to EFTA Court case-law nonetheless manifest clear recognition of the EFTA Court. In view of the ECJ's fairly evident change of opinion about the EEA Agreement in the same period, it appears probable that the EFTA Court's dynamic and integration friendly interpretation has been a contributory factor.
VI. THE CONSEQUENCES: A MORE ‘SUPRANATIONAL’ EEA LAW AND AN UNEXPECTED CONTRIBUTION TO THE DEVELOPMENT OF EU LAW
From the point of view of the EFTA States, part of the price they have to pay for the EFTA Court's dynamic interpretation is a more ‘supranational’ EEA Agreement than originally conceived. As regards State liability, the EFTA Court can even be said to have gone further in Sveinbjörnsdóttir than would probably have been the case had the issue been submitted to the ECJ instead: In light of Opinion 1/91, it seems rather doubtful that the ECJ in 1998 would have derived a principle of State liability from the EEA Agreement. Given that the right of national courts to refer questions of EEA law to the EFTA Court under article 34 SCA was established because the EFTA States, in consideration of their sovereignty, did not wish to open up for preliminary judgments from the ECJ, this is somewhat paradoxical.
Moreover, in some cases, the EFTA Court's interpretation of the EEA Agreement appears to have influenced the underlying EU law. On the factual level, this is clearly the case where EFTA Court judgments trigger subsequent cases in ECJ—for instance when the English Court of Appeal in Boehringer Ingelheim II asked the ECJ whether the EFTA Court's interpretation of the Trademark Directive in Paranova was correct, and when Fokus Bank led the Dutch Gerechtshof te Amsterdam and the German Bundesfinanzhof respectively to refer Amurta and Burda to the ECJ.Footnote 133
Whether the EFTA Court has also influenced EU law on the judicial level is more controversial. However, it seems improbable that the ECJ would refer to EFTA Court case-law purely out of courtesy. It is certainly true that the ECJ refers to EFTA Court case-law primarily in cases where it agrees with the EFTA Court's conclusions, but this is also very much the situation when the ECJ's refers to its own cases. In addition to this, the ECJ's President, Vassilious Skouris, has described the dialogue between the ECJ and the EFTA Court as ‘a paradigm for international cooperation between judicial institutions’ and stated that ‘ignoring EFTA Court precedents would simply be incompatible with the overriding objective of the EEA Agreement which is homogeneity’.Footnote 134 Similarly, the former president of the ECJ's Second Chamber, Christiaan Timmermans, has pointed out that the two EEA courts, through their references to each other are ‘more than paying a courteous salute to the development of each other's case-law’.Footnote 135
A major finding of the present review is that the EFTA Court does not shy away from taking a leading role in the dynamic development of EEA law. However, it should be emphasised that there is no reason to claim that the EFTA Court is in principle a more unreserved supporter of the free market than is the ECJ. When the result nonetheless seems to be that the EFTA Court tends, in hard cases, to follow an even more integration-friendly line than the ECJ, this is presumably mainly due to structural imbalances between the two EEA courts. One should probably be wary of suggesting that the EFTA Court suffers from a ‘little brother complex’, but its case-law leaves one with the inescapable impression that it is striving for recognition from the ECJ. One way of attracting attention would be to take the lead in the development towards further realization of the fundamental freedoms of the common market. The present article shows that it is primarily in such cases that the ECJ and its Advocates General see fit to highlight EFTA Court case-law, a fact which the EFTA Court must surely have noticed.
Further, it is possible that the EFTA Court's desire to prove its independence from the EFTA States has caused it to be even more immune to the arguments of government lawyers than the ECJ. Moreover, one cannot exclude the fear of possible ECJ reactions to imbalances between the Contracting Parties' obligations (remember Polydor and Kupferberg) is a factor leading the EFTA Court to (perhaps unconsciously) insert a certain ‘safety margin’ into its interpretation of EEA law. Should the ECJ subsequently prove to be more sympathetic to the arguments of EU Member States, the only price paid would be that the EFTA States' political latitude would in the meantime have been made subject to greater restrictions than required under EEA law. From the point of view of the EU, this is far preferable to a reverse scenario, in which the EFTA Court would accept national restrictions on the fundamental freedoms of the EEA Agreement which should have been disallowed pursuant to subsequent ECJ case-law. For the EFTA States, on the other hand, the paradoxical consequence is that it appears harder to succeed with arguments for a ‘State-friendly’ solution in the EFTA Court than in the ECJ.
In conclusion, the key question is whether the EFTA Court's dynamic interpretation of the EEA Agreement has been necessary to disprove the 15-year-long rumour of the Agreement's imminent demise. This is a counterfactual question to which there is no clear answer. The above analysis does, however, show that the EFTA Court itself must have considered that a dynamic approach was necessary to convince the ECJ that its earlier evaluation in Opinion 1/91 was far too pessimistic. One may, however, also assume that the EFTA States must have thought that the EEA Agreement would have survived even if the EFTA Court had taken a less dynamic line—they would scarcely have argued for a more traditional international law approach in several of the aforementioned cases had they thought that such a development would jeopardise the whole Agreement.
The EFTA States may well be correct in assuming that the EEA Agreement would have survived even if it had been interpreted to a greater extent as a traditional free trade agreement, but the result would be a wholly different agreement from the one we have today. An evolution of the EEA Agreement in line with the ECJ's predictions in Opinion 1/91 and the EFTA States' arguments before the EFTA Court would have resulted in differences between EEA law and the underlying EU law which would have jeopardized the goal of integrating the EFTA States into the internal market. We must deduce from the present review that a great deal of the credit for the fact that this has not happened is due to the EFTA Court.Footnote 136 Whether the EFTA Court will be able to preserve homogeneity between EU law and EEA law in the future is, however, a very different question.Footnote 137
VII. POSTSCRIPT: THE FUTURE OF THE EFTA COURT
From time to time, the EFTA States complain that the EFTA Court is too dynamic and has gained too much power. Government lawyers regularly argue before Norwegian courts against the wishes of private parties to refer questions of interpretation to the EFTA Court. Further, the right to appear before the ECJ in EEA-related cases is sometimes used to try to undermine unwelcome decisions by the EFTA Court.Footnote 138 Nor have Norwegian authorities shown any interest in extending the EFTA Court's jurisdiction to embrace further aspects of the EFTA States' association with the EU, such as the Schengen Agreement or the new Lugano Convention. It must be said, however, that Norwegian authorities have at the same time loyally adhered to EFTA Court case-law—even where the latter's interpretation of the EEA Agreement is contrary to Norwegian preconditions when the Agreement was signed. Although it is tempting to explain this obedience in terms of Protestant duty ethics, it is probably more correct to view it as pragmatic recognition of the extremely uneven relative strength between the EU and the EFTA States: Norwegian authorities seek to avoid open confrontation with the EFTA Court because of possible reactions from the EU.Footnote 139
From the EU perspective, objections from the EFTA States are a sure sign that the EFTA Court is successfully filling its role as an independent court of justice (see article 108 of the EEA Agreement). Experience of the EFTA Court since 1994 up until the present day is thus a weighty argument in favour of corresponding court solutions in the EU's other association agreements with third countries. The EU seems to have recognised this as, when considering its relationship with Switzerland, it advocated an EEA-style framework agreement with a ‘mechanism for […] homogeneous interpretation of the rules’.Footnote 140
If the ongoing negotiations on Iceland's EU membership should produce an outcome that is acceptable to the Icelandic electorate, the question may arise as to whether an EEA Agreement with only Norway and the ministate of Liechtenstein left in the EFTA camp would justify the continuation of a separate EFTA Court. From Norway's point of view, a ‘Swiss model’ might prima facie seem an alluring option. In this model, Switzerland's own courts are charged with interpreting the country's extensive bilateral agreements with the EU. However, the indications are that Swiss courts operate with a less dynamic approach than the ECJ.Footnote 141 This does not only leave this alternative less attractive from the EU's point of view. Following the reciprocity consideration pursued in this article, it might also lead to a less EEA friendly response from the ECJ.Footnote 142 Given that it seems neither constitutionally nor politically possible for Norway to submit to the jurisdiction of the ECJ, it will hardly be possible to retain the EEA Agreement without the perpetuation of the arrangement with an independent EFTA Court as an guarantor of the EFTA States' fulfilment of their obligations under the Agreement.
