A little over 2 weeks ago, we had an election in Slovenia. The results are in now and there are negotiations going on right now as to who will enter the coalition. And while watching the whole proceeding is quite fun, today I am going to be mostly concentrating on the parties itself.
In Slovenia, one of the organization send all the political parties, that participated in this election (over 20) the questionnaire about their political ideology. So, by answering a couple of questions, each party (at least the ones that answered) were positioned on a couple of continuums:
- whenever human is mostly the result of the genetics or environment
- whenever the criminal system is for punishment or rehabilitation
- whenever we should hold status quo or implement the changes
- whenever internationalism or nationalism is more important
- whenever the property should be in the public or private hands
- whenever the regulation or laissez-faire is better
- whenever the ecology or productivity is more important
- whenever to implement the changes through evolution or revolution
My results can be found both on the picture below and on the following link. The comparison of my answers with different political parties can be found here.
Since most of my classmates (that I discussed this with) voted for the left party (Levica), and I voted for right party (SDS), I had a conversation or two about why I voted the way I did. For myself I knew that I biased myself in favor of the right party beforehand, since I think that I do live in a social country. But there were a couple of things connected to the economy, that I though needed improvement. Since I wanted to vote for somebody, that I was sure that will enter the parliament, I did not have much of the choice (SDS and NSi), and I think the leader of one is more competent than the leader of another.
But I wanted to see, if the results also hold. So, I did a couple of quick analysis, based on the data from the website that I mentioned above. So, I tried to see, which political party in the parliament has the most similar ideology to mine.
I originally wanted to use the k-nearest neighbor, but apparently the sklearn library in python demands labels, so I decided to try something else. But just like k-nearest neighbor, I calculated the distance to each of the party.
Here is the list of the parties in the parliament, from the ones that are most like me, to the ones that are the most different (there is no SDS, since they did not provide the answers):
Depending on the distance used, sometimes the SMC and SNS are exchanged and the DESUS and NSi also exchange places from time to time. But otherwise the order is quite stable. I am quite a right-winger, based on this information.
But to see, if my intuition of being a right-winger is right. Or am I misinterpreting the results. So, what I did is analyze the results of all these parties with hierarchical clustering. Here is one of the results, even though the groupings seem to be quite similar with another parameter used.
Looking at the clusters, there is one cluster, where nobody came in the parliament (Zedinjena Slovenija, Stranka Slovenskega Naroda, Naprej Slovenija), so we can ignore it. On the other hand, there is a cluster of clearly left-wing parties: Levica, SAB, Pirati, SD, Solidarnost and Združena Levica in Sloga. Then there is another cluster of a mixture of right and left, even though most of the parties there are agnostic or right-leaning. There is a right+SMC sub-cluster (SNS, SMC, Kangler and Primic, Rešimo Slovenije Elite in Tajkunov). Then there is another more right-leaning sub-cluster (NSi and GAS), which the only attribute that I see different is, that they seem to have programs concentrating more on the economy. And then there is another sub-cluster, which seems to be concentrated more on helping (LMŠ, DESUS, Zeleni, Za Zdravo Družbo). It is interesting, that this sub-cluster seems to be more connected to right-wing than left, but maybe for one of my future analysis, I could check this.
Now, if I try to see how this connect to which parties I am close to. Let me copy the list again:
They are clearly ordered by clusters. I start with right-leaning parties, then I go through the helping ones, ending up on the left-leaning ones. So, on the end, my bias on the beginning might have be because these parties have more things similar with my thinking than some of the other ones. And maybe next time, I should, because of this, biased myself to the side of the left parties.
The code that I used for the analysis can be found here. The data, that I entered based on the political portret site can be found here.