On 6 April 2016, a referendum on the approval of the EU-Ukraine Association Agreement will be held in the Netherlands. This is the direct result of the Dutch Advisory Referendum Act (DAR), which entered into force on 1 July 2015. According to this Act, citizens can initiate a referendum on most laws and treaties after these have been approved by both chambers of parliament. This referendum may thus be regarded as an experiment with direct democracy giving the Dutch citizens a possibility to refute decisions taken at the political level.
An EU-critical foundation (Burgercomité EU) and a popular anti-establishment blog (GeenStijl) joined forces – under the name GeenPeil – and gathered the necessary 300,000 signatures to call for a referendum on the approval of the EU-Ukraine Association Agreement. That precisely this agreement became the subject of a public debate is a mere coincidence and can only be explained on the basis of timing. It was simply the first legal text approved in the Dutch parliament after the DAR entered into force and, therefore, the first occasion to test its implications in practice. Be that as it may, the crucial question is what the consequences will be when the Dutch citizens reject the Approval Act of the EU-Ukraine Association Agreement. This is not an implausible scenario taking into account that the first polls indicate a significant majority for the ‘no’ camp.
It is striking that none of the official bodies gave a clear-cut answer to this question. Prime Minister Rutte communicated that the government will wait for the outcome of the referendum to decide on its implications. The European Parliament simply took note of the upcoming referendum and ‘trusts that the decision of the Dutch people will be taken on the basis of the merits of the agreement, recognising its tangible effects on the EU and the Netherlands in particular.’ European Commission President Juncker, for his part, warned the Dutch population that a no vote could ‘open the door to a large continental crisis’ without however clarifying why this would be the case.
In principle, the legal implications of a Dutch no vote may be rather limited. The referendum is consultative and there is a turn-out requirement of 30 per cent. However, if the Approval Act is rejected by a significant majority of the population, the political leaders cannot simply ignore the outcome. This also happened after the consultative referendum on the Treaty establishing a Constitution for Europe back in 2005. Despite its consultative nature, the no vote in that referendum implied that the Netherlands was unable to ratify the constitutional treaty. Taking into account the requirements of the EU Treaty amendment procedure (Article 48 TEU), this treaty could therefore simply not enter into force. One may argue that also the EU-Ukraine Association Agreement, as a so-called mixed agreement signed by the EU and its 28 Member States, requires ratification by all parties before it can enter into force. However, there are significant differences between the amendment of EU primary law and the ratification of a mixed agreement. Most significantly, a large part of the EU-Ukraine Association Agreement belongs to the EU’s exclusive competences, either explicitly (such as the trade parts belonging to the Union’s Common Commercial Policy) or implicitly (due to the EU’s implied powers doctrine as codified in Article 3 (2) TFEU). A decision of the Netherlands not to ratify the EU-Ukraine Association Agreement may therefore not have the same far-reaching consequences.
A pragmatic solution to such a (so far hypothetical) situation could be the adoption of a so-called ‘adjusting protocol’ to the agreement. This is precisely what happened after the Swiss Confederation was not in a position to ratify the European Economic Area (EEA) Agreement in the 1990s. The most visible consequence of such a protocol would be the formal amendment of the agreement, implying that “the Netherlands” should be deleted as one of its contracting parties. As a result, those provisions of the Association Agreement belonging to Member State competences would not be applicable in the Netherlands. One could, for instance, think about the provision on mobility of workers (Art. 18 of the Agreement). Taking into account that the most significant parts of the agreement belong to EU competences, the impact of a Dutch non-participation would thus be rather limited. More important, however, will be position of the Netherlands within the Council, which still needs to formally conclude the agreement on behalf of the EU. The European Commission submitted its proposal on 23 May 2013 and the European Parliament already approved the agreement on 16 September 2014, in parallel with the Verkhovna Rada of Ukraine. In accordance with Article 218 (6) TFEU, the Council shall now adopt the decision concluding the agreement. Significantly, such a decision is still pending and has to be adopted by unanimity.
Arguably, a negative outcome in the Dutch referendum should not affect the formal approval of the agreement on behalf of the EU. After all, the scope of the referendum is limited to the question whether the Dutch citizens accept or reject the Approval Act of the EU-Ukraine Association Agreement as adopted in the Dutch Parliament. It, therefore, only concerns the participation of the Netherlands to the agreement. As far as the EU’s participation is concerned, a different ratification procedure applies involving a proposal of the Commission, the consent of the European Parliament and the adoption of a Council decision concluding the agreement. A (hypothetical) Dutch veto to the adoption of the Council’s decision as a result of a negative referendum would neglect the very essence of this distinction. It would not be a victory for democracy as proclaimed by the Dutch initiators of the referendum but rather the opposite. Allowing a relatively small part of the population in a relatively small member state to block the entry into force of an agreement which is approved by the national parliaments of 29 countries and the European Parliament would be very cynical. It would also undermine the consistency and legitimacy of the EU’s external action taking into account that other, largely comparable agreements would remain unaffected.
Finally, a significant part of the Association Agreement already provisionally entered into force. A Dutch decision to reject the approval act will not automatically affect this practice, precisely because the provisional application only concerns those matters falling within the Union’s competence. This caveat is explicitly included in the relevant Council Decisions. In other words, the practical legal implications of the Dutch referendum may be expected to be minimal, even if the ratification of the agreement in the Netherlands would be rejected. The consequences would be more significant at the political level. An overwhelming no vote would be rather embarrassing for the Dutch government at a moment it is holding the rotating EU Council Presidency. It would also be annoying for the EU as such, taking into account the upcoming Brexit referendum and the rise of Euroscepticism on the continent.