Politics and Elections in the Netherlands: Rutte III era

[DavidB.]

The GL win isn't all that impressive if you look at the D66, SP and PvdA losses, which are bigger than GL's gains in almost all municipalities. Even in Amsterdam, where people speak of a left-wing "victory", the losses of SP and PvdA alone are bigger than the gains of GL, not even taking into account D66. Meanwhile, CDA gain a seat and FvD win three while the VVD remain stable, indicating that there must have been D66 -> VVD voter movement too.

However, voter movement from D66 to GL of course does strengthen a more left-wing pole in the system, which might have repercussions for coalition formation and policymaking. In Amsterdam, a government led by the economically illiterate luddites of GL might seriously jeopardize the expansion of Schiphol Airport, which is infuriating.

So who are the biggest winners and losers generally so far?

Winners: local parties, GL, VVD, DENK
Losers: D66, PvdA, SP, PVV, the left as a whole

Some local parties have really impressive results, especially in the Rotterdam metro's (lower) middle-class suburbs. In Barendrecht, Echt voor Barendrecht (EVB) got almost 50% of the vote and won 14 out of 29 seats. Independent New Spijkenisse (ONS) in Nissewaard was already by far the largest party but won an additional seat and now has 14 out of 37 seats. The same goes for Leefbaar Capelle in Capelle aan den IJssel, where Leefbaar Capelle now has 13 out of 33 seats. Inhabitants' Interest Hellevoetsluis in Hellevoetsluis has 12 out of 25 seats. In more middle-class Lansingerland, Leefbaar 3B gained two seats and now has 9 out of 33 seats. Most of these parties (but not all of them) have a profile very similar to that of Leefbaar Rotterdam.

[DavidB.]

Demographic breakdown for the referendum (probably based on the exit poll, so might be slightly skewed in favor of "for") by gender, education and age.

Partisan breakdown (based on GE17).

Percentage "against".


The polls really f**ked this one up; the only poll that came close was Peil.nl's last poll.

This result is disastrous for D66, because of course it will be (almost completely) ignored. VVD, CDA and CU voters won't care, but a lot of D66 voters will, and they take particular interest in privacy issues as well.

[DavidB.]

For those interested: Here the results for the local election by municipality, here for the referendum.

[DavidB.]

This result is disastrous for D66, because of course it will be (almost completely) ignored. VVD, CDA and CU voters won't care, but a lot of D66 voters will, and they take particular interest in privacy issues as well.

How disastrous is disastrous? Will they lose 6-8 seats?

Yes, given D66's record of losing half their seats after governing I'd absolutely take a bet of D66 losing that amount of seats in the next general election (quote this post after the GE to make me look stupid). But just like with the PvdA's loss in 2012, it will probably be not directly due to D66 ignoring and abolishing the referendum but more because of the fact that alternatives (GL, but also VVD) would have a more attractive profile and less negative baggage.

Another consequence of this referendum is that Kajsa Ollongren's popularity has gone completely down the drain; this was already the case before the referendum took place, but if she will be perceived as not sufficiently respecting the result, her reputation will be even more damaged. She was widely considered to be Pechtold's most likely successor (and rumors about Pechtold being done with "The Hague" are getting increasingly loud), but as an alderwoman in Amsterdam she left a mixed impression at best, and she is now among the least popular ministers. Whoever thought that it would be a good idea to put her in charge of Interior Affairs, have her abolish the referendum, and make her push through a law deeply unpopular with a lot of D66 voters was really not quite smart.

[DavidB.]

In The Hague, De Mos ended up with one fewer seat (from 9 to 8 ), the City Party with one more (from 2 to 3). In Rotterdam, hilariously, one seat shifted from the PVV (from 2 to 1) to DENK (from 3 to 4). In Amsterdam, the final result is as follows:

GL 10 (+4)
D66 8 (-6)
VVD 6 (-)
PvdA 5 (-5)
SP 3 (-3)
PvdD 3 (+2)
DENK 3 (+3)
FvD 3 (+3)
CDA 1 (-)
Elderly Party 1 (-)
ChristenUnie 1 (+1)
BIJ1 1 (+1)

So Sylvana has her seat after all, but only if she moves to Amsterdam before next Thursday, otherwise the number 2 on the list will be eligible to take up the BIJ1 seat.

[DavidB.]

The Hague has its excellent website up again, with results at borough, neighborhood and precinct level for both the referendum and the municipal election. Map:


It still feels unreal how I voted with young people, lefties and Muslim voters against conservative, security-oriented right-wingers in the referendum; even weirder because I haven't been in doubt about this vote for a second I also never had the idea that the referendum would be remotely close (except for the hours before the result) and still don't quite grasp what happened and how the polls got it so wrong. I'm so glad I still took the effort to convince less politically oriented friends of mine to vote against, and to have a big poster on my window, despite never even believing in it. Maybe young people really did convince their older family members and friends in the last week. And maybe the editorials in the quality newspapers convinced highly educated older people in the cities. I know the result is going to be ignored again, but it will have a political price and there is now more awareness regarding privacy issues in society.

The government, again, has handled the referendum really clumsily, essentially making exactly the same mistakes as in 2016. One or two days before the election, Rutte managed to make an incredibly disdainful statement: a journalist challenged him for abolishing the referendum, saying that people think it's nice to be able to vote in referendums because it allows them to have a little influence, and Rutte answered that "tv quizzes are also nice, and lacemaking is also nice, but ultimately this [politics] is about serious business."

[DavidB.]

I oppose the law for several reasons.

First, the law would allow the intelligence and security services to share un-analyzed bulk data with foreign intelligence services. I'm down with cooperating with foreign intelligence services if necessary, but we have to know what we share before we share it. There are strict laws for our own government in terms of how long they can save information on telecommunication and internet from individuals (i.e. three years), but our allies may not necessarily have those laws and I think other countries should not have that much information about "non-problematic" Dutch nationals. Moreover, as Americans now know, governments change and today's ally may tomorrow have a very different view of the Netherlands and Dutch nationals.

Second, I am not convinced that we will be any safer under this law. Most terror suspects are already well-known to intelligence services, but the services were too busy concentrating on other potential threats. It makes no sense to increase the amount of information that intelligence services have to analyze. Instead, the intelligence services should have more permissions (and perhaps more resources) to focus on people that are already known to be potential threats.

Third, the law allows the government to store people's DNA and make a DNA bank. I think that's outright creepy and a clear case of government overreach.

Fourth, the government has been lying about why they deem the law necessary, as they have continued to argue throughout the campaign that under the current intelligence law, secret services are not allowed to wiretap communication through the cable, whereas the current law does in fact allow the secret services to do so (which was even acknowledged by the government in a parliamentary debate), though I think they currently need a warrant to do so, which I think is good. It is true, however, that the new law open up a lot more possibilities to collect data in entire neighborhoods just because a few problematic individuals would reside there. I think that is disproportional. By all means wiretap these individuals if necessary (with a warrant from a judge), but not entire neighborhoods.

Fifth, I generally think the erosion of privacy is a big problem, both because of tech companies and because of the government, and this law would have been an additional step back for privacy in the Netherlands. This was a very concrete opportunity to say "no" to this development. I want a government that protects me against companies that want to know everything about me, not a government that treats me as a suspect too.

----
31% of "against" voters were late deciders. Reasons for them to oppose the law:

Biggest reason: privacy, as it should be. Secondary: Sybrand Buma (CDA leader with national conservative profile who might be the biggest supporter of the law), the Cambridge Analytica scandal ("Facebookrel"), "lacemaking" (reference to Rutte's gaffe), the television debate, satirist Arjen Lubach's tv show (Dutch John Oliver and the person who popularized the referendum petition), the arrogant attitude of the government, Mark Rutte, and mistrust of the government.

I think many late deciders who voted "against" were PVV/50Plus voters.

[DavidB.]

The Hague's notorious Schilderswijk was won by the unholy trinity:

Today, mayor Pauline Krikke revealed that she had been robbed there earlier this month while waiting for the tram in the evening. Her bag was stolen. Interesting that we were only allowed to know this after the election.

[DavidB.]

Very good points! I have to say that there was one poll that was very accurate: Peil.nl had it 45-42 with 13% undecided slightly less than a week before the election.


In Emmen, PvdA candidate #12 Ugbaad Kilincci, who has a Somali background, was campaigning on Sunday before the election when some young guys started insulting her, calling her a "black monkey" and "headrag mongol" and saying that she had to "go back to Africa". The incident made it to the national media and sparked much outrage. At the end of the day, however, the PvdA won one seat in Emmen (from 6 to 7) and Kilincci made it to the council with 1007 preference votes, most of which were presumably cast out of sympathy following the incident.

[DavidB.]

YIKES... I feel dirty about agreeing with him!
https://twitter.com/Snowden/status/976953953823920129

[DavidB.]

For GE17 we don't need to rely on Ipsos exits anymore, because we have the NKO. Back then, Ipsos had said that youth turnout was lower than in GE12 (despite total turnout obviously being much higher), De Hond from Peil immediately wrote a sharp rebuttal and turned out to be entirely right based on NKO data. Peil's figures seem credible.

Read the NKO17 and was planning on making a post on that later, highlighting the most interesting findings.

[DavidB.]

smoking behaviour

Is this an actual thing Dutch pollsters ask about in connection to politics?

It's one of De Hond's pet peeves to ask questions about irrelevant things (just like he asks you about owning actual pets). You're likely to find some correlation, but it will always be explained by class, age, education, income or ethnicity, so it is completely useless information. Meanwhile they'll never ask you about class, because that concept isn't supposed to exist in the Netherlands...

[DavidB.]

smoking behaviour

Is this an actual thing Dutch pollsters ask about in connection to politics?

It's one of De Hond's pet peeves to ask questions about irrelevant things (just like he asks you about owning actual pets). You're likely to find some correlation, but it will always be explained by class, age, education, income or ethnicity, so it is completely useless information. Meanwhile they'll never ask you about class, because that concept isn't supposed to exist in the Netherlands...

How would it be measured? Like the British ABC1/C2DE system? There are polls by income (though de Hond never mentions what he defines as "high" or "middle+"), but income may not fully capture things like cultural capital, so results by class could be interesting (for example a moderately successful 50-year old shop owner in Druten vs a 25-year old latte liberal IT consultant in Amsterdam who may have similar incomes but live in 2 totally different worlds).

I suppose self-perception would be used (difficult to measure otherwise, unless you use proxies such as education or income). Part of the reason why this is not being measured is that many people in the Netherlands might not clearly identify as anything in terms of class. How many people in our generation would still identify as working-class, even if they are? Not many, I think. Which is also part of the reason that the left has difficulty attracting working-class voters under the age of, say, 45.

[DavidB.]

Good, here's my NKO post. The NKO is the National Voters' Study, a thorough study in which vote determinants as well as underlying attitudes of the electorate are measured. The study also provides us with valuable raw data on voting patterns among different demographics. The NKO for the General Election in 2017 was recently published; I read it and will here share some of the findings that I found to be the most interesting.

The average age of the parties' electorates. Confirms that the DENK and FvD electorates are relatively young and that the PvdA base is extremely old.


Turnout by age and education: on the left it shows turnout among lower educated people by age group, on the right turnout among higher educated people by age group. The authors suggest there are two main reasons for people to vote: habit and "political confidence". Older, highly educated voters have political confidence and vote because it's a habit. Older, lower educated voters might not have this political confidence, but still vote out of habit. Younger highly educated voters have a lot of political confidence, even though voting is not yet a habit to them. But younger, lower educated voters lack both the political confidence and the habit and therefore many of them did not show up: total turnout was 82%, turnout among lower educated youth only 59%. This, of course, should also be taken into account when analyzing the way young people voted: higher educated people simply make up a much bigger share of the vote among the young than they do among older demographics. Research shows that compared to other EU countries, Dutch youth have a relatively poor understanding of the way our democratic system works. Perhaps improving this understanding would help in increasing political confidence among lower educated young people, because this could become a risk to the "representativeness" of our democracy, especially in times when the traditional "people's parties", once representing "pillars" from the highest to the lowest social classes, are no more and have been replaced by parties representing that only represent smaller segments of society: who, then, will represent lower educated youth?


This figure shows the difference between voters' self-placement on the "cultural" dimension (1= change our culture, 7 = maintain our culture; accidentally mixed up in Dutch) and the way these voters place the party they voted for. Orange indicates the party's (perceived) position, black the voters' position. As other figures show as well, PVV voters consider themselves to be less "radical" than the party. VVD voters, on the other hand, consider themselves to be on the right of their party on the cultural axis, which is why it is a smart strategy for the VVD to comment on "culture war" issues the way they currently do. CU and D66 voters also place themselves markedly to the right of their parties on the cultural axis. It is interesting to note that except for the PVV, all electorates place themselves to the right of their party (though in some cases the difference might not be statistically significant).


The same thing, but then for the economy: 1 = increase income differences, 7 = decrease income differences. The biggest gap is easily between VVD voters and their party's perceived position, indicating that many VVD voters might have voted for them despite considering themselves to be to the left of the party economically. Interesting information for the party, who already campaigned much less on right-wing economic proposals in GE17 than before: perhaps that was a good choice. The very small gaps between the electorates and their parties might indicate that parties' socio-economic left-right position is still a main vote determinant in the Netherlands nowadays, even if this is analysis has nowadays gone somewhat out of fashion. Another interesting point is that all electorates, except for the VVD, place their parties on the "left" (more towards decreasing income inequality than towards increasing income inequality). I personally doubt that's an accurate analysis, especially for "neoliberal" parties like CDA and D66 (but probably also for the PVV and CU).


Same idea: European integration, more (1) or less (7)? Interestingly, D66 voters are much less enthusiastic about European federalism than their party, and the same goes for PvdA voters. Again, except for the PVV, all electorates are less pro-EU than their parties (though in some cases the difference will not be significant). Interesting to note that the VVD electorate is more pro-EU than the electorates of PvdA and CDA, who have been much less critical of European integration than the VVD. How anybody could think that the CDA is not pro-integration is another question... perhaps people believe Pieter Omtzigt.


How did the youth (18-24) vote? Important to take into account this turnout gap among lower educated youth and higher educated youth, meaning that the youth vote skews heavily highly educated, middle-class etc. The yellow thing is the percentage of the vote a party received among the youth, but because of the low N for young voters the margin of error is relatively high, reflected by the long bars. The black thingy is the percentage these parties received in total. Both PvdA and SP were much less popular with the youth (who actually turned out) than with the electorate as a whole; on the other hand, D66 and GL were more popular. No significant difference for the PvdD, dispelling the myth that young women all vote for the PvdD and nobody else does so.


When did people decide on their vote? Options (from left to right): on election day, during the last days, during the last weeks, during the last months, and earlier. Results presented by party. The options "last days" and "last weeks" best reflect the strength of the parties' campaigns. For the PVV, 53% of their voters had already decided to vote PVV months before the election, and relatively few people were convinced by their non-existent campaign, which surprises nobody and confirms that it was almost as if they wanted to lose the election. The option "last days" partly reflects the importance of the diplomatic crisis with Turkey: this seems to have been extremely important to mobilize DENK voters, essentially choosing Turkey's position over the Dutch position by doing so. Very insightful. Other parties with strong campaigns in the last weeks and days were VVD (Turkey) and FvD (thank you, Cambridge Analytica) and, interestingly, 50Plus: Henk Krol had some really strong debate performances. GL (Klaver effect), CDA (Buma's nationalism) and D66 (why?) also managed to attract quite some people in the later stages of the campaign. Also note just how small parties' real "base", i.e. people who will always only vote for one party, has become.


The NKO also confirmed that most people decide between a small number of ideologically similar parties: the idea of voters who are completely clueless and decide between random parties is baseless.

[DavidB.]

Great post, David!

The figures are perhaps too low to be significant, but I thought the SGP voters were having lots of children, so their young voter share would be a bit higher, but that seems not to be the case.

Thank you. I expected this too. However, the N is probably really low, so I'm not sure if I'm ready to draw any conclusions, but at least these findings do call into question this idea.

I think I have an ok perception of most parties and their core voters, but I'm a bit doubtful on CU. How would you draw up a stereotype of a typical CU voter. Of course fairly religious, but what in terms of education, sector of work, (dis)satisfied with politics etc.? From the graph, it seems their average voter is slightly younger than average.

Have to be honest here: I have to rely on what we know in political science, because I know only a handful unrepresentative CU voters; most "bubbles" or spheres in Dutch society I understand, but this one is difficult for me to "get" as well.

I would think of an average CU voter as a fourty-something married woman in the East of the country who works in education or healthcare, has an average education background, is relatively religious (maybe slightly to the right of the mainline option within the Protestant Church of the Netherlands), cares about the environment, is worried about certain changes (growing intolerance on all sides of the political spectrum, climate change, EU integration, losing our Dutch norms and values, increasing number of lonely elderly people) but at the same time feels blessed living in a wealthy country, is not dissatisfied with politics and does not worry too much about making ends meet. People close to her may vote CU, but also CDA or VVD.

CU also does well within certain black evangelical communities ("Hallelujah churches") in Amsterdam Southeast; I imagine that the gender gap would be even bigger there than among white Dutch people. The CU councilman who was just elected in Amsterdam is black.

[DavidB.]

Clickable map for all precincts in Rotterdam both for the local election and for the referendum. Compare here for GE17. The Leefbaar vote held up really well in areas that were strongly VVD in the GE. For instance, in the northernmost polling station in newly built middle-class Nesselande, the VVD got 39% in the GE; D66 and PVV came second and third with 13%. In the local election, 31% voted Leefbaar here, 22% VVD and 12% D66 (and 2% PVV, what a joke). DENK also improved their score really well in "their" neighborhoods, especially when taking into account NIDA taking part as well (who also got impressive percentages in these areas).

Of course, turnout in Rotterdam was 46% in this election and 75% or so in the GE.

[DavidB.]

Is there a lot of NIMBYism when it comes to red light districts and where brothels operate?

Yeah. The VVD support the "Dutch model" nationally but want to ban or heavily restrict prostitution almost everywhere locally for NIMBY reasons (the same goes for "coffee shops" where pot is sold). There are no prostitution areas in Rotterdam and Utrecht anymore. The prostitution area in The Hague turns the whole neighborhood to sh**t. At almost every time of the day, you will see "customers" as well as shady Bulgarians and Poles in huge cars with young hot women next to them. It leads to a decrease in housing prices and a concentration of poverty and social problems.

As mvd10 said, PvdA and SP are shifting towards the Swedish model on prostitution (though they are not there yet) for feminist reasons: prostitution is capitalist exploitation of vulnerable women, and very few people would choose to be a prostitute if a lack of money were not an issue. PvdD seem to take a similar position while maintaining "soft" support for the Dutch model.

GL are split on this issue, but they will never go for the Swedish model. SJWs tend to oppose that model because they get their ideas from America. This is a pretty big issue for BIJ1.

[DavidB.]

Deadlock in the coalition formation in Rotterdam, where D66 and PvdA block Leefbaar's attempts to engage in coalition talks with them. The VVD do want to govern with Leefbaar. In The Hague, talks are led by former VVD leader Hans Wiegel, who will attempt to form a coalition consisting of De Mos, VVD, D66 and GL first. In Amsterdam, GL aim at forming an oversized coalition consisting of GL, D66, PvdA and SP.

Edit: PvdA, GL and D66 will now try and form a coalition in Rotterdam. With NIDA, DENK and SP, they would have a majority, but the optics of cooperating with NIDA and especially DENK would... not be too great.

[DavidB.]

Are local PVV groups under the same cordon sanitaire as the national party?

National parties leave this to their local branches, so no, not really. However, at this point it seems unlikely that the PVV will govern anywhere, though I would not rule it out altogether in places like Spijkenisse, Venlo or Rucphen. In almost all places (Spijkenisse and Almere seem to be the exceptions) the PVV is no less extreme on the local level than it is on the national level, so the ideological distance between the PVV and other parties is simply too large. What's more, governing with the PVV will lead to a lot of national media attention and scrutiny, which isn't exactly something local parties want to have to deal with. And then there's the issue of scandals. You don't want to be in a coalition with "Rutte should be hanged" (Rucphen), "Germania heritage" (Zoetermeer), and "Burn the mosques" (Utrecht) types of people (the ones in Zoetermeer and Utrecht were elected), and it's likely more scandals with PVV councilmembers will be dug up over the course of the next four years. The PVV train has truly derailed (I've not seriously considered voting for them in the local election for a second) and it's both sad and entertaining to see what will happen next.

[DavidB.]

Based on the referendum result, the government has decided to amend the Law on Security and Intelligence Services by introducing a clause into the law stating that data cannot be intercepted randomly: interception has to be focused on specific targeted individuals. What's more, before bulk data can be shared with foreign intelligence services, the intelligence services will have to show that the services with which information is shared adhere to a similar understanding of democracy and the rule of law as in the Netherlands. More specifics will follow, but this does sound good.

[DavidB.]

In The Hague, the first round of negotiations between De Mos, VVD, D66 and GL has, somewhat surprisingly, been successful. Former Public Health Minister Edith Schippers (VVD) will now lead the next round of coalition talks. The five priorities of the four parties will be sustainability, drawing up an agenda for the inevitable growth of the city, "livability" (related to safety/security), "everyone participates", and mobility/traffic.

[DavidB.]

The formation process in Rotterdam is in deadlock, as the VVD currently refuse to engage in talks with PvdA, GL and D66 without Leefbaar. Now, FvD have brilliantly proposed to officially end their alliance when (and only when) Leefbaar start governing with D66 if this would be necessary to make a coalition between Leefbaar and D66 possible. Brilliant, because breaking up the alliance does not actually mean anything (it is obvious that neither the very small ideological distance nor the warm feelings between Leefbaar and FvD have changed), but FvD/Leefbaar get to look constructive and D66 lose their most important argument to decline cooperation with Leefbaar. I would absolutely love for D66 to be cucked into a coalition with Leefbaar and VVD.

[DavidB.]

The government has gotten into trouble over the already unpopular policy of scrapping the dividend tax, which will cost 1.4 billion euros annually. This policy was not in any party's manifesto, so people were surprised when it was suddenly part of the coalition agreement and the question was: who pushed for this policy to be implemented, and what are the benefits? We unofficially know that Rutte himself put forward this policy and we unofficially also know that Shell and Unilever requested for this policy to be adopted, but when asked whether there had been a memo by the Ministry of Economic Affairs with pros and cons, both Rutte and Eric Wiebes (VVD, Economic Affairs) had said no. This already seemed highly unlikely, as politicians just don't decide on a 1.4 billion policy without a memo by the Ministry of Economic Affairs or the Ministry of Finance, and now University of Amsterdam researchers found out through a request based on the Freedom of Information Act that hey, there was indeed a memo. Another big fat lie by Mark Rutte, but the government are trying to spin it as if the memo was not a memo but something else. The coalition parties have each other's back on this, but it does not look good and this is an amazing opportunity for the opposition, and particularly for GL and PvdA who have been going on about this all the time, to paint the government as only interested in the interests of multinationals (and it's hard to deny they have a point...); big business lobby organization VNO/NCW's proposal to double the property tax on ordinary households to make up for the costs does not make all of this look any better, to say the least...

[DavidB.]

NRC Handelsblad pointed out that this is the third time in one year that Rutte got off well, but never managed to convince the opposition or the public that he told the truth: first came the Security and Justice affair (this feels much longer ago though...), then Halbe Zijlstra's hilarious and catastrophical Putin lie, and now this; in all three cases, Rutte's position was questionable. Insiders say Rutte is considering running for a fourth term after this one, and I guess he would win another election (and even be the best potential VVD candidate), but at some point stuff like this has to erode his position, you'd say...

Another interesting point was that D66 and CU leaders Pechtold and Segers came to Rutte's defense, but CDA leader Buma refused to do so, saying he was an MP and not responsible for the government's handling of this issue. A sound position from a constitutional perspective, but he handled it in a very clumsy way and ended up looking worse than anyone else.

[DavidB.]

Completely in agreement with mvd10's analysis.

Today's Peil poll has quite some shifts compared to last week due to the debate on the abolition of the dividend tax, with the government parties all losing seats:

VVD 24 (-1)
GL 17 (+1)
FvD 16 (+1)
SP 15 (+1)
CDA 13 (-2)
PVV 13 (+1)
D66 12 (-1)
PvdA 12 (+1)
PvdD 8 (nc)
50Plus 7 (nc)
DENK 5 (nc)
SGP 3 (nc)

Only 18% now support the abolition of the dividend tax, compared to 26% last week. 40% of GE17 VVD voters, 17% of GE17 CDA voters and 19% of GE17 D66 voters support it. 76% think Rutte lied, 73% think Buma, Segers and Pechtold lied, and 61% would have voted for the motion of disapproval that was supported by all the opposition parties except for the SGP. 67% think Rutte "often has difficulty with the truth", and 71% think Rutte should not remain PM after this term. A rather damning picture.