Flags of the Dutch E.U. presidency and campaign posters for the nonbinding referendum on the E.U.-Ukraine association agreement wave in The Hague on April 6. (Peter Dejong/AP)

On April 6, the Dutch voted against accepting the association agreement between the European Union and Ukraine. Roughly one-third of the Dutch electorate went to the polls; of those voting, 61 percent voted no.

This is the second time that the Dutch have voted no in an E.U.-related referendum. In 2005, a similar majority rejected the European constitution, with nearly two-thirds of the electorate going to the polls and 61.5 percent voting no.

Technically, this agreement was not about Ukraine joining the E.U. — the agreement was mainly, though not exclusively, a trade deal. The referendum’s implications for the association agreement are unclear, but some observers are already claiming that this referendum was about more than the association agreement. To them, the size of the April no vote indicated a rejection of the European Union.

What should we make of this second Dutch no vote? What are its implications?

What was the referendum really about?

Strictly speaking, this referendum was about an association agreement between the E.U. and Ukraine, which first and foremost would have made trade easier between the two. The largest chunk of the document deals with establishing a “Deep and Comprehensive Free Trade Area,” although other sections require Ukraine to adhere to “democratic principles, rule of law, good governance, human rights and fundamental freedoms.”

However, one of the Dutch anti-association agreement campaign organizers, Arjan van Dixhoorn said, “We don’t really care about Ukraine.” The goal was to put pressure on the Dutch relationship with the E.U.

In July 2015, the Netherlands became one of a handful of established democracies to enact a statute allowing citizens to trigger a referendum on a law or treaty that is approved by the parliament, before it becomes active. The association agreement was simply Dutch citizens’ first opportunity to force the government to hold a referendum about an E.U.-related topic.

This was also the first opportunity to test the referendum legislation itself. The statute includes two elements that are in conflict. First, any referendum’s results are nonbinding — which means that the government can ignore what the voters have said. Second, there is a turnout quorum of 30 percent, which means the vote is only valid if at least that portion of the electorate goes to the polls.

But what does that mean, in practice? Did the turnout quorum play a role in the minds of voters? This referendum met the quorum, making the voters’ rejection of the association agreement valid. How does that comport with the first clause, which says it’s nonbinding?

Who voted what, where and … when?

The no camp won by a wide margin: 61.00 percent to 38.21 percent. Though the turnout was low, it passed the 30 percent floor. An overwhelming majority of the districts voted against the association agreement, although some areas in the big cities, such as Amsterdam and Utrecht, voted in favor of it.


Dutch E.U.–Ukraine referendum of April 6. Municipalities in red voted no. (Wikimedia Commons)

What does this outcome mean? Polls and analyses carried out immediately after the referendum pointed to the following:

  • Highly educated voters tended to vote yes.
  • However, a lot of highly educated voters did not show up.
  • To some extent, this reflects a failure of the campaign of the yes camp: It struggled to establish a full-scale get-out-to-vote effort.
  • The turnout quorum seemed to have played a crucial role in depressing both the turnout and the yes vote: A lot of the yes voters went to polling booths just before they closed or remained at home. I will explain this paradox below.

Voter’s remorse

The turnout quorum created a conflict for the yes campaign and its supporters. If they stayed home, the result might be invalidated. However, it would also increase the percentage of no-voters. Conversely, voting could have the opposite effect: It would reduce the percentage of no-voters, but it could make the outcome valid. Simulations on the night before the referendum showed that, given the polls, for every 1 percent drop in turnout because of strategic vote abstention, the no portion of the vote would increase by about 2 percent.

Since the turnout was close to the quorum’s line, many pro-agreement citizens said afterwards that they felt remorse — both for voting and for failing to. According to a poll carried out by Eenvandaag, 20 percent of the yes-voters regretted turning out to vote. Conversely, 14 percent of the non-voters regretted not turning out to vote.

If all of the remorseful yes-voters had abstained, the turnout would have been just below the 30 percent quorum, with a two-thirds majority voting against the agreement. Conversely, if all remorseful yes-leaning non-voters had voted, the turnout would have been around the 40 percent mark — with only a tiny majority voting against the agreement.

We can draw three lessons from this:

  1. The yes camp faced a severe coordination problem. Earlier research found that national referendums in E.U. countries had a markedly lower turnout when there was a turnout quorum. The Dutch referendum shows that indeed a turnout quorum offers voters strategic incentives not to vote and thereby depresses the turnout.
  2. The ‘No’ majority was inflated by strategic abstention. The polling data suggests that a sizable part of the pro-association citizenry did not vote in an attempt to deflate the turnout and invalidate the result.
  3. Both of those facts suggest that the referendum result doesn’t simply indicate Euroskepticism. The 2005 Dutch referendum clearly displayed dissatisfaction with the extent and form of European integration at the time. That’s not so clear in this referendum. Some Dutch citizens do seem to fear Ukraine joining the European Union. But many other factors mattered as well.

What did we learn from the referendum?

We don’t know yet exactly what will happen to the association agreement. But the vote will probably affect both domestic and international politics in several ways.

First, the Dutch referendum legislation must be taken seriously into account. If Dutch governments are negotiating E.U.-related treaties, they must be ready to be called to defend them at the polls. Anti-E.U. civic forces are now experienced at running and winning a referendum campaign.

Second, the Dutch won’t necessarily vote against European integration every time, but clearly the government will have to get better at campaigning and coordinating strategy to avoid another disastrous outcome where the turnout is right above the turnout threshold and the no vote is severely inflated.

Third, theoretically Dutch referendums are nonbinding, but in practice it will be hard to ignore the outcomes, especially given the turnout quorum.

Fourth, and not surprisingly, the government wants to eliminate the quorum requirement. It’s not clear whether parliament would vote to repeal it.

Finally, E.U. observers wonder whether the Dutch result will influence other such votes. Nigel Farage, the leader of the British U.K. Independence Party, recently claimed, “If the Dutch vote decisively in a no direction, it will have an impact on the U.K. referendum [to leave the E.U.].”

This is unlikely. The Dutch no may (slightly) affect the Brexit-campaign discourse, but, contrary to what some have claimed, it won’t necessarily encourage the British to vote to leave the European Union. After all, the outcome was heavily influenced by the turnout quorum. The Dutch referendum signals many things, but a rising tide of Euroskepticism is not one of them.

Kristof Jacobs is an assistant professor in comparative politics at Radboud University, where he studies democracy, electoral systems and referendums and blogs for the Dutch political science blog www.stukroodvlees.nl.