Web-Based Assistant Changes the Face of Dutch Politics

An anonymous reader writes "The elections held in The Netherlands on Wednesday have shaken the country. Almost 10 million votes were cast, and statistics show that a full half of those who voted used a popular web-based voter guide. This guide is operated by the independent institute for the public and politics. Advice is given to the visitor upon answering a number of multiple choice questions on some common political topics. Statistically, a number of people ended up scoring in support of populist parties both on the far left and far right. No bias was reported to exist in the test itself. However, these parties have ended up with an unforeseen amount of power as a result of the election. The voter participation was high, and the web-based advisories may have motivated people with little interest in politics to cast a vote anyway. Can politics be simplified to a ten minute test?"

more info

[Ubi_NL]

It should be noted that this voting aid is endorsed by all major political parties who actually submit questions to it. The party leaders are also the first to take the test (this time the liberal leader actually did not end up with his own party at all after doing it...bummer :(

In the end you can compare your answer to the one of each political party. There they argue why they give this answer, making it a rather nice tool to learn more about the programs without reading the entire manuscripts, but it is definitely more then just the 30 questions.

Another interesting thing is that there is no large correlation between the suggested votes and those actually casted, indicating that people do not follow the advise blindly. In reality, many people here try a number of these web-based aids (kieskompas.nl is another one).

Voting Compass

[Killjoy_NL]

I tried Stemwijzer, but the questions where too simple with only yes and no as possible answers.

I tried Kieskompas.nl and they had better questions, followup questions and at the end you could compare your "score" with that of the political parties that answered the same questions accompanied by extra explanations and motivations to give you a better idea about their standing on the subjects.
That was a better website to "quickly" get informed.

Revealing what was always there?

[Myself]

If it opens people's eyes to parties outside the usual two, I'm in favor of it. Play with the OkCupid politics test [okcupid.com] if you haven't already. It's run by the same mathematicians who designed TheSpark way back when, and features the same scarily-insightful ratings engine.

Re:I took the test

[malsdavis]

Seriously, the way it works is just pure propaganda. It reduces choices to black or white.

umm, that isn't anything like what propaganda is.

From m-w.com:
1 capitalized : a congregation of the Roman curia having jurisdiction over missionary territories and related institutions
2 : the spreading of ideas, information, or rumor for the purpose of helping or injuring an institution, a cause, or a person
3 : ideas, facts, or allegations spread deliberately to further one's cause or to damage an opposing cause; also : a public action having such an effect

How does "It reduces choices to black or white" satisfy any of those definitions?

Re:Interesting

[diamondmagic]

In simple terms:
You are not voting online... The program is recommending you a party to choose when you do go and vote.
We are also dicussing the option of making the actual ballot like this as well, with the ability to recommend a party.

Re:Interesting

[mabinogi]

it's not an online voting system, it's a recommendation system. There's no reason at all why something similar couldn't be used in the USA.
IP Addresses have nothing to do with it.

I'm voting for Fortuyn!

[0jjjjjjjjjj0]

Well, apparently. After being asked the 30 questions, they suggest I vote for Fortuyn, which translates to what, exactly?

Here are the 30 topics, each of which you are asked to 'agree' with or 'disagree' with.

• 1 Citizens should elect the prime minister.
• 2 Child benefit should be increased.
• 3 The more you earn the larger the contribution you should pay for health insurance.
• 4 Working parents should pay for child care facilities.
• 5 If you need home care you should contribute towards it.
• 6 Young people under the age of 27 should no longer receive social security benefits.
• 7 Everyone receives state old age pension from the age of 65. People who in the future receive state old-age pension and a pension of over 15,000 euro should themselves pay contributions towards this benefit.
• 8 It should be easier for employers to terminate the employment of staff with a fixed contract.
• 9 Cannabis and other soft drugs should be decriminalised.
• 10 Tackling terrorism is more important than the personal freedom and privacy of the individual.
• 11 Everyone over the age of 14 is required to be able to prove who is he or she is. The government should abolish this identification obligation.
• 12 Everyone should be free to say what he wants, even if this discriminates against other people.
• 13 A teacher at a school with many children who have learning difficulties should earn more than a teacher at an ordinary school.
• 14 Junior general secondary schools (MAVO) no longer exist as a school in their own right but have become part of the preparatory secondary vocational education system (VMBO). The MAVO should be restored as a separate type of school.
• 15 Christian and Muslim faith schools should have the right to refuse pupils.
• 16 The government currently subsidises the public television channels Netherlands 1, 2 and 3. One of these channels should be axed.
• 17 Euthanasia should be allowed.
• 18 Everyone should make clear during their lifetime whether their organs may be donated to sick people after their death.
• 19 Some people have a Dutch passport and a passport of another country. The government should abolish this dual nationality.
• 20 Antilleans who commit crimes should be sent back to the Antilles and imprisoned there.
• 21 A homeowner receives tax relief on mortgage interest. This scheme should be made less advantageous for people in the higher income groups who take out a new mortgage.
• 22 Landlords of residential property should be able to decide for themselves how much rent they will ask.
• 23 No new mosques may be built.
• 24 Schiphol Airport should continue growing.
• 25 More roads are needed to combat traffic congestion.
• 26 The Netherlands should abandon nuclear energy in due course.
• 27 The rights of animals should be safeguarded in the Constitution.
• 28 Turkey should be able to join the European Union.
• 29 The Netherlands should spend more money on defence.
• 30 The Netherlands should no longer participate in the Joint Strike Fighter (JSF) programme.

Freedom of choice

[asciimonster]

Well, this election in the Netherlands some people concerned with the abovementioned effect (viz. a newspaper and a university) have created http://www4.kieskompas.nl/ [kieskompas.nl]a competing site(unfortunately no english verions available) which wanted to provide a more graduated result. Hell, there was even http://www.partijwijzer.nl/ [partijwijzer.nl]a similar website(currently offline for obvious reasons, i.e. elections are over) aimed at younger (age < 30) voters.

As long as there are more than one what-should-I-vote websites and most people visit several, I don't see the problem.

Re:Interesting

[Anonymous Coward]

It is written in javascript, and is running locally in your browser. Go ahead and audit it.

Re:Statistically

[KimV]

Just to make it clear, the pedophiles' party couldn't run. A party needs 570 pledges of support from voters (30 in each of the 19 districts) in order to take part in elections and they failed this requirement.

Re:Too easy to create bias

[Anonymous Coward]

"It's all about who gets to edit the questions..."

With all due respect, I didn't see these questions in the survey. Were you meaning to be rhetorical or make a pointed example?

I design testing and survey instruments for my employer, and there is a LARGE scientific basis behind all of it. You look for biases on your own. You bring in content experts from a sampling of the population. You run test surveys on a random sample representing the population that will be utilizing this. You run item analysis on all of these. Stats alone will find bias, even if you are a complete idiot. Generally, this information is freely available to representative groups (naturally, in the idea of testing, it is limited, but they work from the same idea).

To go over all the work that goes into something like this, you'd probably need to have several stats courses behind you, tests and measures courses, a few psychometric courses etc. It isn't something people get into lightly.

Its nice to be able to sum this stuff up with a few extremely biased questions that would never be approved by a bipartisan group and would really be more representative of a one-sided push poll (i.e., the bias is obvious, but you want to prime the poll-taker to take your side next time this subject comes up). And this is why any REAL survey group that was doing something like this would state their backers, those that have signed off, and the background. In the case of the question above, my group would ask for a short response from both sides and guide them towards language neutral wording (some words such as incompetent would be ruled out immediately without even assigning sides...luckily, I haven't had to work on public policy type debates, but knowing those that have, these groundrules are put forth early on and STILL argued and refined by the sides).

I wish it was all this simple, someone creates a bullshit poll and then we sit around collecting money and ignoring the guys that state that its all about who gets to edit the questions, but we don't.

I wish I could post un-anonymously, but part of my job is being politically neutral and even commenting on this sort of stuff publicly could ruin future work. Whats the phrase I'm always reminded of -- its not impropriety that folks have a problem with, its the appearance of impropriety. Or something like that.

Re:Other factors

[AlXtreme]

The strong shift to both extremes was indicated by polls before the introduction of the many online tests. As a result, we can conclude that the online tests didn't have a significant influence.

The reason for this shift is simply because large parts of the population aren't happy with the current government.

Re:Too easy to create bias

[StikyPad]

The questions were very straightforward. "Should landlords be allowed to decide their own rent?" "Should people under 27 continue to receive social security?" "Should animal rights be included in the constitution?" Even the trickiest hot-button question (in the US, at least) was about as neutral as possible: "Should tackling the terrorism problem take priority over individual freedoms and liberties?"

Re:Too easy to create bias

[Hershmire]

survive the flood of imported goods?

It also helps to read the question properly.