In an important judgment delivered on Wednesday September 9^th, Ferreira da Silva and others (C-160/14), the European Court of Justice (‘ECJ’) for the first time reprimanded a national supreme court for failing to start a preliminary reference procedure. On the same day, it also ruled that the Dutch supreme court was not obliged to refer a question when a lower court had already done so in X & Van Dijk (C-72/14 and C-197/14). Read in tandem, the cases shine an interesting light on the acte clair doctrine, which determines when courts of last instance are freed from their obligation to refer a question to the ECJ.

The acte clair doctrine, first articulated in Cilfit (C-283/81), states that in accordance with article 267 TFEU, a national court against whose judgment there is no higher appeal must refer a question to the ECJ when a question on the interpretation of EU law arises, unless the answer is “so obvious as to leave no scope for any reasonable doubt”. It is part of a set of rules that intends to create a full system of legal remedies in the EU and protect legal unity. The rule seems clear enough, but it leaves courts of last instance with significant discretion. After all, whether something is beyond reasonable doubt is principally for the national court to decide. The ECJ itself drastically narrowed the scope of the principle with the requirement that the matter must be equally clear to the courts of other Member States, taking into account the different language versions of EU legislation (there are now 24 official languages in the EU). This requirement, when taken literally, is so stringent as to provide little practical guidance.

Ferreira da Silva concerns the interpretation of the phrase “transfer of a business” in the sense of the Transfers of Undertakings Directive 2001/23. Several former employees of an airline operator sought to rely on the directive after the company was wound up. The Portuguese supreme court refused to refer a question to the ECJ, reasoning that the interpretation of the phrase was beyond any reasonable doubt. The applicants then started a new case, claiming damages from the supreme court for failing to refer a question when it was obliged to do so (a possibility introduced by the ECJ in Köbler). At this point, the national court decided to refer a question to the ECJ, asking among other things if the supreme court was under an obligation to refer a question to the ECJ, considering that the contested phrase had been explained differently by lower courts in Portugal and elsewhere.

The ECJ starts modestly by emphasizing that national courts have the ‘sole responsibility’ to decide whether a matter of EU law is beyond reasonable doubt. The fact that other courts have given contradictory decisions is not decisive in triggering the obligation to refer a question. However, the phrase in question has given rise to ‘a great deal of uncertainty’, not just in Portugal but also in many other Member States. This, according to the ECJ, creates a ‘risk of divergences in judicial decisions’ within the EU. In that case, the acte clair doctrine does not hold, and the court of last instance is under an obligation to refer a question.

Meanwhile in X and Van Dijk, the Gerechtshof (court of appeal) in Den Bosch asked preliminary questions on the application of the social security provisions of Regulation 1408/71 to Rhine boatmen falling under the international Rhine Agreement. The Dutch supreme court, in a similar case, reasoned that the answer to those questions was so obvious that it needed no clarification. It then started a preliminary reference to ask whether it had to await the outcome of the pending questions of the lower court, or that it could dismiss the question as an acte clair. Here the Court was more lenient. The fact that a lower court has referred a question to the ECJ, does not in itself preclude a supreme court from concluding, in line with the Cilfit criteria, that the case is an acte clair.

The two cases should not be seen as a departure from the Cilfit criteria. The ECJ has repeatedly confirmed the doctrine, despite criticism on its practical application. Still, without explicitly saying so, the Court seems to have slightly slackened the strict requirements of Cilfit. Assessing the risk of divergences in judicial decisions, domestically and elsewhere, seems more feasible than consulting all language versions of EU law from the perspective of every single court. X and Van Dijk demonstrates that courts have a certain freedom to assess this, and that a mere contradictory judgment of a lower court does not preclude the finding of an acte clair. On the other hand Ferreira da Silva shows that the ECJ is also willing to step in where a court of last instance has obviously misapplied the principle. This more workable approach seems beneficial for the relationship between national courts and the ECJ and the smooth functioning of the preliminary reference procedure.