Since 1 July 2015, the Netherlands has been one of the few Western democracies that has allowed its citizens to call a referendum. However, the legislation allowing referendums is now repealed, as the new government says the law has not delivered what was expected. Four serious referendum attempts have been made during the life of the provision, of which two resulted in actual referendums. Based on these four experiences, Kristof Jacobs highlights what other countries can learn from the Dutch referendum experience.

The Dutch referendum law

The Consultative Referendum Law allows citizens to trigger a non-binding referendum about a law or treaty within a strict timeframe. In a first phase citizens have four weeks after the law or treaty is approved to collect 10,000 signatures. When this threshold is met, citizens then have six weeks to collect 300,000 new signatures. Certain topics, such as the constitution, the budget, and the monarchy are exempt.

Once enough signatures have been collected, a referendum is held, and 30% of the electorate must take part in order for the vote to be valid. If the majority of voters support ‘No’ and the turnout threshold is met, a bill withdrawing the law or treaty has to be introduced. The Parliament can decide not to approve this bill, but a debate has to be held. In other words: the result is advisory, even though there are formal consequences. In practice, both successful referendums led to changes to the existing legislation/treaty.

The referendum law does not allow government-initiated referendums, for two reasons. First, if the referendum is supposed to be a method of highlighting gaps between the government and public opinion, it is widely accepted among politicians that they are not in a good position to be the ones calling the referendum. After all, if they already knew there was a gap, they would surely already have acted to close it. Moreover, such referendums are seen as easy to manipulate by government parties.  

Table 1. Overview of the main referendums and referendum attempts

Topic	Signatures 1st stage	Signatures 2nd stage	Outcome
Association treaty Ukraine & EU	13,480	427,939	Turnout: 32.28% No: 61.00% Yes: 38.21%
Power of intelligence services	17,162	384,126	Turnout: 51.54% No: 49.44% Yes: 46.53%
Homeownership tax deduction	24,822	11,306	NA
Organ donor legislation	19,266	146.636	NA

The first referendum: populists testing the water

Eurosceptic organisations started to look for suitable legislation to have a referendum on the EU as soon as the referendum law was implemented. The Association Agreement between the EU and Ukraine was the first opportunity to test the waters.  

On 10 July 2015 the Eurosceptic organisation Burgercomité-EU and the shock-blog Geenstijl joined forces and announced that they would start a campaign to collect signatures to trigger a referendum. A second organisation, Forum voor Democratie joined them and they succeeded in collecting enough signatures, despite the holiday season. In part this success was due to a strong online and ground campaign and (especially) ample media attention.

The referendum was held on 6 April 2016 and the ‘No’ side won.  The topic was complex and the initiators of the referendum openly admitted that they were treating it as a referendum about the democratic legitimacy of the EU, rather than about the Association Agreement. As a result, the referendum campaign dealt with many different issues.

The official National Referendum Survey 2016 (NRO) later revealed, however, that the voters were not particularly interested in rejecting the EU. Dutch citizens did seem to fear Ukraine joining the European Union, but many other factors, such as corruption in Ukraine and the desire to support Ukrainian democracy mattered as well.

In the aftermath of the referendum, the government tried to ease the worries of the No voters by negotiating an addendum to the Association Agreement. However, as the 2017 national election survey later revealed, voters were not convinced and roughly 64% of the respondents indicated that the government should have withdrawn its signature from the Agreement.  

The second referendum: post-materialists strike back

After the Ukraine referendum, several attempts to pass the first hurdle of 10,000 signatures failed. Only in the summer of 2017 did a group of university students manage to get enough traction to progress to the second round of signature collection. The topic was very different from the first referendum; it was about the powers of the Dutch intelligence services. The students claimed that new legislation would violate the right to privacy and give the intelligence services too much power.

The students were supported by a host of civil society organisations, such as Bits of Freedom and Amnesty International. However, they only managed to cross the threshold after they got the support of the progressive comedian Arjen Lubach, presenter of popular late-night TV show Zondag met Lubach.

This time the topic could be boiled down to a simple choice between two values: privacy versus national security. As a result, the campaign debates were more focused. Most importantly, both sides agreed that some form of updated new legislation was needed. Even the ‘No’ side made it clear that they did not want the entire bill to be thrown away.

While polls predicted a win for the ‘Yes’ side, they also noted that potential ‘Yes’ voters were less informed and that people who knew about the topic were more likely to switch to the ‘No’ side. In the end, more people voted ‘No’ and the government conceded defeat.

The two smallest parties in the coalition government, the Social Liberal Party (D66) and the Christian Centre Party (ChristenUnie) pressured the two larger government parties into changing the legislation. Several changes were made that were in line with the demands of its opponents. While some ‘No’ campaigners were unhappy with the outcome, polls suggest that the majority of Dutch voters were content with the changes.

The failed referendum attempts

After the second referendum, two more attempts were made to trigger a referendum, but both failed to pass the 300,000 threshold.

The first attempt was one by the party for the elderly, 50 Plus, who wanted a referendum stopping the phasing out of the tax advantage for homeowners who repaid their full mortgage. This legislation would disproportionally hit the elderly and the party wanted to raise its profile and mobilise its base. However, it seriously underestimated the efforts it takes to collect signatures. While it got the support of elderly organisations, it failed to devise a campaign structure and only managed to collect 11,306 signatures.

The second attempt was made by Geenstijl, who wanted a referendum about the new organ donor legislation. The new law would reverse the existing opt-in policy into an opt-out one, in the hope to increase the number of organ donors. The law was very divisive, and was only approved after one MP allegedly missed his train connection.

Geenstijl also wanted to use the referendum initiative to signal their discontent with the intended abolition of the referendum legislation. Polls showed that the abolition of that legislation was deeply unpopular; its rapid approval would make a referendum about the organ donor law impossible. Geenstijl was hoping to get more media attention because of this ‘race against the clock’. However, the Dutch parliament indicated it would not prevent a referendum on the donor law, leading to a loss of media interest in the referendum attempt. The campaign, which largely existed only on social media, collected 146.636 signatures and failed.

 Lessons & the future

The Netherlands has now abolished its advisory, citizen-initiated referendum legislation. What can we learn from the experience?

Undeniably, the first referendum on Ukraine left a ‘bad taste’. Many highly educated Dutch people (who overwhelmingly voted in favor of the Association Agreement) felt the referendum was hijacked by populists. But the second referendum made clear that the referendum is an instrument that can be used by groups across the political spectrum – just as elections can be used by populists and non-populists alike to gain influence. In that sense the referendum is more of a neutral tool than a populist weapon, one that can be used for better or for worse.

Opponents of the referendum also felt that the signature threshold was too low, that in times of social media it was too easy to collect the necessary signatures and that only populists would be able to trigger a referendum. The next three major referendum attempts proved these claims wrong.  

Indeed, at the moment it is clear that the thresholds seem appropriate, that social media campaigning alone is not enough to trigger a referendum and that the referendum can also be used to vote on topics that post-materialist and educated citizens care about. As such it is ironic that the government wanted to abolish the legislation because it poses a populist threat.

This irony did not escape a special State Committee on democratic reform. Even though the committee is chaired by Johan Remkes, a member of the party most opposed to the referendum law (Prime Minister Mark Rutte’s VVD), its report advised in favour of referendums. In fact, it even called for stricter, binding referendum to be allowed. The issue of referendums is likely to stay on the political agenda for quite some time.

About the author

Kristof Jacobs is Assistant Professor in Comparative Politics at the Political Science Department of Radboud University. He mainly studies referendums, populism, democratic reform, and social media. He is currently finishing a special issue for Acta Politica about the Dutch referendum experience and was the project leader of the Dutch 2016 and 2018 national referendum surveys.