Blog of the International Journal of Constitutional Law

Going Against the Consociational Grain: The Debate on the Dutch Advisory Referendum Act and the Ukraine-EU Association Agreement Referendum

Reijer Passchier & Wim Voermans, Leiden University

On April 6, 2016, the Netherlands held the first referendum under its new Advisory Referendum Act of 2015 (Wet raadgevend referendum)[1] and the third national referendum in two centuries.[2] This was a test on a highly controversial issue: Dutch political (consocialist) culture and referendums do not mix well.

The referendum was on the question: ‘Are you for or against the Approval Act of the Association Agreement between the European Union and Ukraine?’ Though the turnout was a meager 32.28% (just reaching the 30% turnout threshold for a valid referendum), a clear majority of 61% of the voters checked the box ‘against’. Although the outcome merely presents ‘advice’ to the government, it was felt that the ‘no-vote’ needed to be followed up upon. That means that according to the Advisory Referendum Act (ARA), the original ratification Act for the EU Association Agreement, that had already passed both Houses of Parliament, now needs to be wheeled back. Under the ARA the Dutch government must, if it it wants to annul the original ratification act, propose a follow-up Bill to repeal the ratification act. The other option is to neglect the outcome of the referendum and to decide to put the original ratification act into effect, and thereby abide by Parliament’s will. Dutch referendums are corrective in nature – with acts of parliament as its subjects – but they are not legally binding. This corrective, but yet advisory non-binding nature of referendums takes some getting used to in Dutch political culture. On the one end of the political spectrum, some say that the government cannot ignore what basically amounts to a ‘vote of no confidence’ against (further) European integration. On the other end, commentators deny the relevance of the referendum, highlighting the fact that all-in-all a ‘mere’ 19,7% of the total electorate cast a vote against.

Indeed, the question how to follow-up on the April referendum is a major headache for the Dutch government who currently preside the European Union.

The Ukraine Treaty has 28 signatories and 27 member states have already ratified. With the present outcome of the referendum, which is already hard to read for Dutch commentators (is it an undirected outcry of Eurocriticism, frustration over the EU’s migration crisis, lingering political dissatisfaction with the present policies of the coalition, distrust of the Ukrainians, or just and uniformed emotional mishmash?), it is utterly incomprehensible for the European partners and the Ukraine. The major issue is of course how to translate this outcome into a clear position for renegotiation, if any. Currently the Dutch stand totally isolated with a terrible negotiation position. Out of serious options, the Dutch government could therefore do not much more than it did: postpone its decision on the outcome.

Joyful feelings about the success of the referendum are therefore lukewarm at best. Most Dutch commentators feel that this first experience of holding an advisory referendum under the new Advisory Referendum Act was not an altogether gratifying experience. This blog reports on three of the most prominent critical claims made. Although we think that some of these claims do make sense to a certain extent, we will argue that the greater part of the discontent with the referendum actually goes a lot deeper: referendums seem to go against the grain of the time honoured Dutch political consensus culture.


Any referendum under the Advisory Referendum Act starts with an introductory request requiring at least 10,000 signatures from enfranchised citizens. If this requirement is met, the government is, under Art. 2 of ARA, obliged to organise a referendum if the introductory request is followed-up by a definitive request requiring the endorsement of at least 300,000 enfranchised voters. This indeed seemed to be a formidable hurdle to take in a country with approximately 12.5 million eligible voters at the time the ARA act was devised back in 2008. No one at the time, however, could have fathomed how easy it is to collect signatures with electronic devices and clever apps. It took the initiators of the Ukraine referendum only days to collect enough preliminary signatures and the use of dedicated apps made the collection of the 300,000 signatures for the definitive request more or less a walk through the park. Voters were very easily mobilized for the initiative: with a click of the finger they could sign up, as they did during more or less casually in the streets, at happenings or on trendy websites. Although this was clearly not the ease of signing up the architects of the ARA had envisaged, the Dutch Elections Council (Kiesraad) could not do anything else than decide that these electronically harvested signatures were valid.

Lots of commentators were disgruntled with this and argued that the ARA makes it way too easy for individuals and groups to launch referendums. The 10,000-300,000 signature threshold proved to be a laughing matter in this day and age of social media and internet, or so this argument goes. The referendum instrument thereby can be too easily be hijacked and manipulated.

Depending on the question whether one believes that a referendum is per definition a ‘celebration of democracy’ or an instrument that should not be too readily available, this claim may make sense to a certain extent. However, we believe that part of the criticism delivered on the signature requirement is also prompted by a deeper discontent concerning the topic of the referendum on the Association Agreement. Some of the commentators who have brought the signature requirement up for discussion also claim that Eurosceptical parties did indeed hijack the referendum instrument to use it to reveal popular discontent with European integration in general. Whether or not this claim holds true, we cannot really oversee. But it did confront the Dutch with the sobering fact that politics surrounding a referendum is just as ambiguous as politics in Parliament.

The 30% Turnout Requirement

Article 3 of the Advisory Referendum Act provides a quite peculiar requirement for a non-binding advisory referendum; under a minimum turnout of 30% the government does not need to consider the outcome of the referendum. Falling short of 30% turnout requirement renders a referendum invalid. This somewhat strange requirement (a sort of a third hurdle) raised a lot of eyebrows, especially because it turned into a central element during the Ukraine-Treaty referendum. At the outset of the April referendum, in March, the polls already made it pretty clear that a majority would cast a negative vote, but it was uncertain whether the turnout requirement for a valid referendum would be met. After the referendum, many commentators complained that the 30% turnout requirement impaled those in favour of the Act on the horns of a difficult dilemma: vote ‘for’ – and risk contributing to rendering the no-vote majority outcome valid; or not turn up to vote – and risk the turnout requirement being met anyway, lose out on your vote in favor, thereby helping the no-vote-camp to an even greater majority. Moreover, a lot of the critics claimed that the combination of a turnout threshold and the non-binding character of the referendum does not make any sense at all: the turnout requirement is both confusing and useless. At best it wrongfully suggests that the referendum has a binding character after all.

In the eye of the public meeting, the 30% turnout requirement indeed felt like decisive for the victory leaving no options to the government than to comply to the Dutch voters will, as indeed, according to large volumes of Dutch referendum voters, the rest of the EU Member State that undersigned the Treaty should. In the Dutch media it came as a bit of a shock that the EU did not automatically bend to the will of the Dutch voters. The outcome met with a chill comment of the European Commission on behalf of the 27 Members States that had indeed ratified the Treaty: the negative outcome is a problem for the Dutch and for them alone. The EU Commission as early as 7 April made it clear that the “no” in the referendum on the Association agreement between the EU and Ukraine, has no effect on the agreement itself. The agreement was already signed and unanimously adopted by all EU Member States’ heads of government, and the Dutch “no” does not change that. The Dutch government therefore is effectively cornered on the domestic and EU-level: no room for renegotiation of the Treaty and no other option than to bend to the will of the no-voters and revoke the original act ratifying the Treaty. The Rutte-cabinet tried to avert the heat of the debate to by some time, which is difficult enough because the ARA requires a proposal for revocation of the Act subjected to a referendum ‘as soon as possible’. Campaigners from the no-vote camp were infuriated by the delay the Rutte-cabinet and have taken it to court (with only the flimsiest chance of success, by the way, because the court will undoubtedly consider it to be a political question and therefore not rule on it). The government is in dire straits indeed due to the failing setup of the Referendum Act. The advisory character of the referendum gets overcast by the political realities of the day: 30% turnout requirements, together with a total lack of experience of the Dutch with referendums, trigger false expectations that are difficult to counter and were effectively not countered at all at the occasion of the first referendum under the Act.

We feel, like a lot of commentators, that the Advisory Referendum Act should be modified here by doing away with the 30% turnout requirement, and reconsidering the volumes and nature of signatures for initial requests and endorsements further down the road. However rooting for change in the current Referendum Act at this moment in time, with the Act only having been used once, is predominantly perceived as panicky reactions of poor loosers.

A Referendum about Treaties?

A last bone of contention concerns the possibility of referendums on Treaties under the Dutch Referendum act. Some believe that the system of the Referendum Act is way too liberal in this respect. Article 4 of the Dutch Referendum Act provides that referenda can be held on Acts ratifying a treaty and even on tacit approvals of such Treaties.[3] The system of the Referendum Act allows for a wide range of Acts eligible for a referendum. Article 5 of the act only excludes (a) Acts regarding the Monarchy; (b) Acts regarding the Royal Family; (c) Acts regarding the state budget; (d) Acts regarding the revision of the Constitution[4]; (e) Acts which exclusively implement Treaties or decisions of treaty organisations; and (f) statutes that regulate matters for the entire Kingdom (rijkswetten). Acts ratifying Treaties are not excluded and, interestingly, the possibility of excluding such Acts was not even considered when the Advisory Referendum Bill was discussed in Parliament. Some claim that the legislature has overlooked this issue and that it was a serious mistake to permit referenda on Treaties. The current setup of the Advisory Referendum Act can make the Netherlands an unreliable treaty partner because, even after the government has signed a Treaty and Parliament has ratified it, a small proportion of the voters (with a 30% threshold a meager 16% of the total electorate can win the day) might force the government to hold a referendum before the act ratifying the Treaty can enter into force.

Although there is some merit to this argument, we do not expect substantial negative effects for Dutch diplomacy in this respect. Indeed, the Netherlands is not the only country that can organize referendums on the ratification of Treaties – it is a quite common phenomenon. In some presidential systems, it is even common for presidents to sign a Treaty without them ever being sure beforehand that their parliament or Congress will subsequently ratify it. Considering that such uncertainties do not commonly cause insurmountable problems in foreign affairs either, we expect that the same will go for the new Dutch system for Treaty approval. And even then, the Advisory Referendum Act will enable the people to advise the government on a variety of matters so why should the Dutch be so afraid of allowing the people to bring about non-binding recommendations regarding matters of primary European and international law?


The April 6 referendum surprised the Dutch as much as it did their European counterparts and Ukraine. Even though it was not a big domestic or European issue, it caused upheaval, and raised as many questions as eyebrows. What was the matter in this tiny country with its longstanding open trading tradition and its international orientation as a result thereof? Was it a burst of propped up and lingering resentment against elites, domestic and European alike? Was it political underestimation and lack of experience with referendums of the political class? Was it the result of a flawed Referendum Act that silently slipped into the Dutch constitutional system in the summer of 2015, under the radar, virtually unnoticed?

The single cause for the shock it caused is difficult to tell right now. One plausible explanation for the magnitude of the shock is the head to head collision of a referendum, as a binary instrument of direct democracy, with the constitutional and political tradition of the Netherlands. The Dutch political system is a prime example of a consociational political system, in which deep internal religious and ethnic divides are bridged by an ‘inclusive political system’ that is characterised by various forms of division of powers, proportional representation and the protection of minorities.[5] This system values negotiation, compromise, and consensus – there is even a special name for it: ‘poldercultuur’. The electoral system of absolute proportional representation, with no form of electoral threshold, both at national and decentral level, is in line with this at least a hundred year old consociational tradition of the Dutch system. This explains why the Netherlands, until recently, was a more or less referendum-averse country and maybe – in view of recent events and the critical voices raised – still is. Referendums were debated for more than a century in the Netherlands, with no success until 2015. There is something in the very character of the Dutch constitutional and political system that is at odds with this ‘blunt’ instrument of a referendum.[6]

Yet, the Dutch now have their referendum Act, and more referenda will undoubtedly follow. Time will tell whether we will ever get used to it.

Suggested Citation: Reijer Passchier & Wim Voermans, Going Against the Consociational Grain: The Debate on the Dutch Advisory Referendum Act and the Ukraine-EU Association Agreement Referendum, Int’l J. Const. L. Blog, June 1, 2016, at:

[1] This is the second Act passed on referenda. The first Act was the Temporary Referendum Act of 2004, which had facilitated the referendum on the Treaty on a Constitution for Europe.

[2] The referendum on the Ukraine-EU association agreement was the third national referendum in more two centuries. In 2005 a referendum was held on the Treaty on a Constitution for Europe and way back in 1798 the Constitution for the Batavian People was put up for a referendum as well. In both referendums the proposals were rejected.

[3] According to the Act on the Ratification and Proclamation of Treaties (Rijkswet goedkeuring en bekendmaking verdragen), the Dutch legislature can either approve Treaties by adopting a Ratification Act after a debate on a bill to that extent, as was the case with the Ukraine-EU Association Agreement, or by tacit consent (i.e. without a debate) to a proposal.

[4] It may be interesting for constitutional scholars to know that Acts regarding the revision of the Constitution are excluded from the possibility of holding an advisory referendum, because the legislature considered that introducing such a referendum by way of ordinary legislation would be unconstitutional. The explanatory memorandum to the Advisory Referendum Act (par. 7.2) argues that the formal revision procedure of the Dutch Constitution (Article 137) provides an obligation for the legislature to, if it has been passed in first reading, consider a Bill to revise the Constitution in second reading. Moreover, according to the explanatory memorandum, an advisory referendum on constitutional revision under the Advisory Referendum Act would violate Article 139 of the Constitution which stipulates that: ‘Amendments to the Constitution passed by the States General and ratified by the King shall enter into force immediately after they have been published’.

[5] John McGarry & Brendan O’Leary, ‘Introduction: The macro-political regulation of ethnic conflict’. In: John McGarry & Brendan O’Leary (eds.), The Politics of Ethnic Conflict Regulation: Case Studies of Protracted Ethnic Conflicts. London: Routledge 1993, p. 1–40.

[6] See: Voermans, Wim J. M., A Bird’s Eye View of Dutch Constitutional Law (May 11, 2016), par. 2. Available at SSRN:


Leave a Reply

Your email address will not be published. Required fields are marked *