By Benedikt Pirker
All eyes were on the Wightman case in recent days. This may have somewhat overshadowed a second interesting development: On Friday 7 December the Swiss government (the Federal Council) decided to publish the result of its negotiations with the European Union on a Framework Agreement (FA) for their bilateral relationship. Such an agreement would form a sort of governing structure for the most important of the Bilateral Agreements that currently link the EU and Switzerland.
There is a complex political context to the negotiations of this draft agreement that I will deliberately leave aside for the present post (see for a recent overview over Swiss-EU relations here). To put it in a nutshell, since 2008 the EU requests this step from Switzerland, and since 2014 the EU and Switzerland have been negotiating a special agreement to cover the most crucial current and future (market-access oriented) agreements among the Bilateral Agreements currently in force between Switzerland and the EU. The goal is to create a more reliable framework (1) for Switzerland’s incorporation of EU legal acts in the relevant domains, (2) for the uniform interpretation and application of the Agreements and the EU law referenced therein, (3) for the surveillance of the application of those norms and (4) for the settlement of disputes (Article 1 (3) FA). Presently, I want to highlight two elements that seem to be of relevance beyond the confines of Swiss-EU relations: the solution found for the interpretation and dispute settlement of the FA and the law it covers.
The political context
It should be noted that there will be – at least from a current perspective – a difficult and protracted political struggle for the FA to find acceptance on the internal political plane in Switzerland. The publication of the draft for the purpose of a national consultation of stakeholders is widely seen as an attempt to gain time in this context. The Swiss negotiators achieved some notable parts of what they wanted (see a nuanced assessment here). Nonetheless, the general political climate within Switzerland is predominantly hostile to a number of aspects of the FA. For example, measures taken to protect the local high wages in Switzerland have been a concern for the EU, whereas they are seen as indispensable on the Swiss side. The resulting compromise in the FA is likely to cause discontent in Switzerland on various sides. There are a number of other issues specific to the Swiss-EU context (e.g. the adoption of State aid rules or of the Union Citizenship Directive by Switzerland), but it would lead too far to expand on them here (there are good overviews in English here and in French from the official Swiss perspective here).
Some observers have quickly gone as far as to judge the situation as hopeless and argue that Switzerland should abandon the very idea of a framework agreement, although such an agreement is indispensable to continue with the current system of the Bilateral Agreements. Instead, avenues such as accession to the European Economic Area or a dock-on solution to the latter should be explored in their view (see here).
I think, however, that despite the unpredictable fate of the draft FA some interesting lessons can and should be learned from it. Notably, a look at its dispute settlement provisions and rules on interpretation should be taken. If those provisions should serve as a guidance for the future, it becomes clear that any long term solution to participate “from the outside” in the EU internal market entails considerable power given to the CJEU. Moreover, it entails considerable restrictions for domestic courts with regard to the interpretation of any “parallel” norms to EU law contained in an agreement with the EU. The alternative seems to be the European Economic Area (EEA) – which, however, has its own drawbacks. In many cases (such as the Swiss one) the EEA may offer an alternative version of dispute settlement and interpretive rules, but at the price of accepting the rest of the EEA “package”. This cure may then be worse than the original illness, if it implies accepting other norms that would be politically even less acceptable (in Switzerland, think e.g. of many aspects of free movement of persons law currently not covered under the Bilateral Agreements). Let us now turn to the two topics in turn.
Dispute settlement in the FA
The FA rules on dispute settlement generally follow typical standards of state-to-state arbitration in public international law, with the exception of disputes where the interpretation of EU law is at stake. Notably, the chosen framework for dispute settlement is exclusive for both parties concerning interpretative disputes (Article 9).
To briefly sketch the procedure, if the relevant joint committee cannot find a solution to a dispute, after three months the EU or Switzerland can ask for the establishment of an arbitral tribunal. If the dispute raises a question of interpretation or application of one of the norms of the FA, the covered agreements or a referenced EU legal act (as provided for in Article 4 (2) FA), the arbitral tribunal turns to the CJEU, if the interpretation of that norm is relevant to resolve the dispute and necessary to enable the tribunal to take a decision (Article 10 (3) FA). The judgment of the Court binds the tribunal. If after the tribunal’s decision a party considers that the other party has not complied with the decision, it can take compensation measures up to a suspension of part of or a whole agreement to remedy a potential imbalance. The proportionality of such measures can be brought before another arbitral tribunal if the joint committee does not take a decision on the matter within six months (all of this in Article 10 FA).
The most notable aspect in this article is the strong formulation of the duty of the Tribunal to bring matters before the CJEU (“le tribunal arbitral saisit la Cour de justice de l’Union européenne”, emphasis added). Just for comparison, Article 174 (1) of the Draft Withdrawal Agreement for the United Kingdom similarly requires that the arbitration panel “shall not decide” any question of interpretation of a concept of EU law and that it “shall request” the Court to give a ruling on the question. The provision also contains a possibility for a party to request the panel to review its assessment (Article 174 (2)). Nonetheless, apart from this review element the two approaches appear broadly comparable. It seems that an approach for negotiating uniform dispute settlement mechanisms with partner countries is emerging on the EU side.
Now, one might read such norms and be enticed to think about the leeway for the FA arbitral tribunal to avoid requesting a ruling from the CJEU. If we read Article 10 (3) FA as a strong command, the tribunal has little leeway in deciding to reject a party’s argument that it should turn to the CJEU. The Tribunal essentially becomes a somewhat pointless exercise that simulates autonomous dispute settlement, but actually serves as a way station for the parties to the CJEU. By contrast, the tribunal may behave like some EU Member State courts in the context of the preliminary ruling procedure and try to exploit the leeway it has. Article 10 (3), after all, speaks of “questions” that must arise, an interpretation that has to be “relevant” (pertinente) to resolve the dispute, and “necessary” (nécessaire) for the Tribunal to decide. At each juncture, at least theoretically it appears possible that the Tribunal would try to carve out some interpretive leeway for itself. However, there is a catch. The more leeway we interpret into the norms on the Tribunal, the less palatable the FA’s arbitral tribunal solution appears from the perspective of the CJEU.
Already in Opinions 1/91 and 1/92 that led to the creation of the EFTA Court, the Court was confronted with a similar problem. The first version of the EEA Agreement was to create a court that would be competent to interpret a common set of norms parallel to EU law and that would be competent both for the EEA EFTA states and the EU and its Member States. This “EEA Court” was also to take binding decisions on the interpretation of the relevant norms. The CJEU insisted that there could not be such a court taking decisions on this parallel body of norms that would be binding on the CJEU itself. Moreover, it required that the court to be created instead of the EEA Court (the EFTA Court) was only an institution created between the EEA EFTA states, and that the CJEU could not be obliged to pay due account to its decisions.
It can already be seen from this that the Arbitral Tribunal in the FA follows these red lines of the CJEU in its design. It would interpret parallel norms to EU law, and provide interpretations binding for both the EU and Switzerland. Consequently, there has to be an obligation for the Tribunal to submit questions on EU law to the CJEU for a binding decision by the latter. Therefore, our small thought experiment on the leeway of the Tribunal quickly meets its limits.
In a nutshell, if we can speak of a new model of dispute settlement in external agreements like the FA, their design possibilities are quite constrained. The FA tribunal system is extremely closely bound to the CJEU, which in turn makes it politically very difficult to sell (in this case in Switzerland). A very far-fetched reading favouring the Arbitral Tribunal’s leeway of decision-making would be unacceptable for the CJEU. This apparently inevitable dispute settlement structure may be less of a practical concern in the Brexit context, where such dispute settlements provisions will be less relevant in future relationship treaties that are expected to make fewer references to EU law. In the Swiss context, however, the FA with this strong role for the CJEU is supposed to become the framework of Swiss-EU relations for many years.
The interpretation of the FA and the law it covers
Next to the rules on dispute settlement, the FA also contains rules on interpretation. The preamble of the FA already points out that the parties are conscious of the need for an institutional framework ensuring “homogeneity” in the parts of the internal market in which Switzerland participates (consideration n° 8). Article 4 contains the “principle of uniform interpretation”. To reach the goals of the FA, respectful of “the principles of public international law”, the covered agreements and referenced EU legal acts are to be interpreted and applied in a uniform manner for all the parts of the internal market in which Switzerland participates (para 1). As far as notions of EU law are concerned, the FA, the covered agreements and the referenced EU legal acts are to be interpreted and applied in conformity with the case law of the CJEU. This concerns case law handed down before, but also after the signature of the “relevant agreement” (para 2).
This is nothing completely new. Two of the earlier Bilateral Agreements, the Agreement on the Free Movement of Persons and the Agreement on Air Transport, already contain provisions that require to follow interpretations handed down by the CJEU for concepts parallel to EU law before the signature of the relevant agreement. However, for later case law they only require that it be taken into account and discussed in the Joint Committee. The Swiss Federal Tribunal has found a rather pragmatic strategy to deal with this different treatment for older and newer case law. In the framework of the Agreement on the Free Movement of Persons, it expanded its willingness to follow the case law of the CJEU also to jurisprudence handed down after the date of signature of the agreement, holding that it would do so generally unless there were “pertinent grounds” to develop an alternative legal construction of a norm. As a consequence, the Swiss interpretation of the Bilateral Agreements has been broadly aligned to the CJEU’s approach, with the Federal Tribunal keeping a safety valve ready (that has not been used to date). One may also be inclined to compare this approach to the situation for homogeneity in the European Economic Area (EEA). The principle of homogeneity in Article 6 of the European Economic Area Agreement read together with Article 3 (1) and (2) of the Surveillance and Court Agreement means that EEA norms parallel to EU law must be interpreted in conformity with past CJEU case law, whereas later case law must only be paid “due account”. In practice, the EFTA Court has similarly decided to follow the CJEU also with regard to later case law (see e.g a remarkable revirement de jurisprudence on these grounds), while keeping the door open to deviate from such case law if there are convincing reasons (see, e.g., here for one example of a different legal context).
It constitutes a strong statement in this context – and perhaps implicitly a signal of lack of trust, given the track record of the Federal Tribunal and the other Swiss courts that so far seems to be rather respectful of EU law and the CJEU’s interpretations – that the FA now contains a clause requiring homogeneous interpretation with regard to past and future case law. In a nutshell, dynamic homogeneity is required, without any visible possibilities for Swiss courts to at least formally maintain the possibility to deviate from CJEU case law under special circumstances.
While understandable from the point of view of the EU’s constitutional system, this clause is undoubtedly problematic for pretty much any domestic legal (constitutional) order. As the EFTA Court puts it in Jabbi, “[w]ithout independence in its adjudication no court could claim legitimacy” (para 71). Under the discussed clause, a state has to agree that its courts be bound unconditionally to the future interpretations given to provisions of an agreement by another court, namely the CJEU. Here, however, we are not in the situation of an EU Member State court embedded in the judicial structure of the EU, including the cooperation with the CJEU in the framework of the preliminary ruling procedure. Neither are we in the situation of the EEA EFTA states whose courts may request an advisory opinion from the EFTA Court (which is not formally binding under Article 34 Surveillance and Court Agreement, but given considerable weight in practice). Moreover, in the EEA context, as shown above homogeneity is read by the EFTA Court (and certain EEA EFTA state courts) as less apodictic than it seems to be enshrined in the provisions of the FA. In other words, there seems to be no wiggle room for diverging interpretations by national courts under the FA, not even the largely theoretical leeway claimed by the EFTA Court and national courts of EEA EFTA states in the framework of “EEA-style homogeneity”.
What is to be learned from the FA?
The Swiss situation and thus the FA could be read as being simply a very specific solution in a very specific context. The whole conundrum is also at least partly caused by aspects of path-dependency of the Swiss-EU negotiations. The Federal Council put the possibility of settling interpretive disputes with a role for the CJEU in the Swiss negotiating mandate in late 2013. Perhaps, as a consequence, it was to some extent inevitable to get to the point of a draft FA as it has been presented above. And it should be noted in this regard again that in several respects, the Swiss side has managed to achieve remarkable concessions that are enshrined in the FA, which render it probably the best result possible given the limiting circumstances.
Nonetheless, there are also broader lessons to be learned on the evolution of dispute settlement and interpretive provisions in such agreements. Switzerland may to a considerable degree simply be “at the wrong place at the wrong time” and therefore be among the first to be confronted with these provisions in the FA. It appears to be clear now that based on the constitutional structure of the EU, the negotiating strategy on the side of the EU – in the Swiss context as much as in the Brexit context – is that a very strict form of uniform interpretation with EU law is the price to pay for any sort of internal market participation “from the outside”. The CJEU’s case law requires to ensure the Court’s comprehensive interpretive authority over EU law in this context. The discussed provisions on interpretation and dispute settlement are the consequence. This means that in the future, there seem to be only two options for internal market participation “from the outside”. On the one hand, option A that the FA’s provisions set out entails a considerable loss of domestic judicial interpretive independence in matters covered by the relevant agreement; on the other hand, option B consists in joining the European Economic Area and learning to live with the latter’s advantages and drawbacks.