Note to reader: I highly suggest that readers read the text describing the findings before looking at the animation. This paper does not follow the common academic principles and was made to practice data wrangling and data visualization. Due to the limitation of text and overflow of animations, I will only make short analyses of the findings.

Introduction

In this project I will try to analyse how the politics and popularity of political parties in 14 EU-countries have varied over time. I will both look at individual party data but also aggregated data on country and ideological level. To do this, I will analyse data from the Ray-Marks-Steenberger and Chapel Hill expert survey. Most focus will be on the two variables, the modern GAL-TAN scale for social and cultural values, and the classical LEFT-RIGHT scale for economic policies. Together they make up the “political compass” which is a more modern and suitable way of analyzing political parties. I will also look at popularity of different ideologies over time and changes in GAL-TAN orientation for the included countries.

Research questions

To analyse and understand the European and Swedish political landscape over time I will try to answer the following research questions:

  1. Has the popularity of the political ideologies changed?
  2. In what way has the Swedish parties changed their values over time and does it differ from the rest of Europe?
  3. Could we see any regional patterns in GAL-TAN orientation?

The first question will look at overall popularity trends of the different ideologies between 1984-2019. The second question will take a deeper look at the variables GAL-TAN and LEFT-RIGHT for the Sweden specific case and compare it to aggregated data for all European parties from the years 1999-2019. The third will analyse movement in only GAL-TAN at country level from 1999-2019. Also, I do not aim to describe why in this project, I will mostly look at what has happened.

Data

For this analysis I have used data from two different data sets, the Ray-Marks-Steenberger survey and the Chapel Hill Expert Survey. Both data sets covers a broad range of variables about political parties in EU to map the political landscape. The data comes from experts in the field, mostly economists and political scientists, that have analysed political parties and tried to quantify their policy proposition and ideological stand points on a scale between 0-10. Every party is categorized with ideological family belonging. The first datasets cover 1984-1999 and the other from 1999-2019. Only the second one has values for GAL-TAN, LEFT-RIGHT.

Table 1: Variables and description
Name Description
Family/Ideology Ideological belonging
Country Country
Year Year of measurement
Party Party abbrevation
Vote Percentage vote in latest election
LRECON Measure of economic policies where 0 means far left(total ownership by the state) and 10 far right(ultra free market capitalism)
GALTAN Measure of social and culture values where 0 is green, alternative and liberal values(GAL) and 10 is traditional, authoritarian and nationalistic (TAN)

All data can be found on their websites [https://www.chesdata.eu/ray-marks-steenbergen-survey , https://www.lsu.edu/faculty/lray2/data/data.htmln]

Analysis

Has the popularity of the political ideologies changed?

In graph 1 I have animated the popularity of different ideologies from 1984 until 2019. The aggregated data was created by calculating mean votes of all parties in a specific family/ideology in a given year. What we could tell from this graph is that the conservative and the social democratic parties, for example Moderaterna and Socialdemokraterna in Sweden, have always been popular in EU but in the latest 10 years we have seen a loss in popularity. The voters have changed relatively evenly to the other parties making the political landscape more equal between ideologies.

The other findings from these graphs includes the increase in parties in the “no party” family, from 2014 to 2019. Many of these parties focus only on one topic and therefore could not be classified due to their simplicity and abscence of policy for other topics. Also, we could see a rise of the green parties compared to data from 1984, since they start from a low level the increase must be seen in percentage, not percentage point. Christian parties and agrarian seem to have loosed over time and radical right parties gained votes. Another interesting result is that radical right parties have not gained as much as one would think from looking at how media and the general public portrays this issue.

Graph 1: Popularity of ideologies in EU from 1984-2019

In what way has the Swedish parties changed their values over time and does it differ from the rest of Europe?

To answer this question, I will look at the political compass which combines GAL-TAN (Y-axis) and LEFT-RIGHT (X-axis) in order to position political parties more correctly. Further up on Y-axis indicates more traditional, authoritarian, and nationalistic (TAN) values and further down indicates more green, alternative, and liberal values(GAL). I will call this social and cultural values and typical questions that influence this variable is immigration, political freedom and law and punishment. On the X-axis we have economic policies, where further to the right indicates a more positive attitude against markets, free trade, low tax etc, and further to the left indicates bigger government interventions, welfare, higher tax.

Although this graph shows voting percentage (votes for KD, MP and SD look unproportional big as they have two letters in the abbreviation), we will focus on the movement of the parties. I have also chosen not to show the parties Piratpartiet and Feministiskt Initiativ as they are irrelevant for this analysis due to their small impact. From 1999 until 2019 we could see that Centerpartiet (C) has made the biggest change from a more traditional (TAN) farmers party to a (GAL) liberal party. The other parties have been relatively steady during the years which contradict common arguments that states that Socialdemokraterna (S) have abandoned their social democratic values.

The most interesting part about this animation is what happens when Sverigedemokraterna (SD) joins in 2010. SD is the most TAN party due to their anti-immigration policies. First, we could see that the other parties, except Kristdemokraterna (KD), try to take distance from SD by moving more to GAL to make the distance for voters bigger, i.e bigger political leap to take. When the parties notices that the voters still leave for SD they instead take on the strategy to move closer, more TAN, in order to win them back from SD.

Graph 2: Swedens political compass from 1999-2019

To see if Sweden is different to the other EU-members we will compare graph 2 with graph 3, which shows us the average location of the parties from the 14 EU-members grouped by family belonging. One finding is that Sweden does not have any regional parties in the parliament, other countries like Spain and Belgium have parties representing different regions like Basque and Flanders. Also we could see that our Agrarian party, Centerpartiet, is more aligned to GAL values then the rest of EU the parties.

The overall results shows that the political landscape is quite similar but with some minor differences. In Sweden the majority of the parties are more GAL than their EU counterparts, most noticeable with the radical left, liberal and agrarian parties. As in the Swedish case many parties seem to move more to TAN when the radical right parties gain influence (could be seen when the label, Radical Right, grows).

Graph 3: EU’s political compass from 1999-2019, grouped by family

Could we see any regional patters in GAL-TAN?

GAL-TAN could be viewed as social and cultural values for a country and has lately been used much by media and in academia. In this graph I will look at how the analysed countries have switched in their GAL-TAN orientation from 1999. This value is also weighted by the popularity of the parties, i.e bigger parties will influence the country more and therefore contributes more to the value. A higher value means more TAN oriented and a deeper blue color on the graph.

From 1999 until 2010 the general trend has been that countries have moved more against GAL, indicating greener, more liberal values. After 2010 some countries have moved more against TAN values.

Sweden is a clear example as well as France and UK. When colors shifts it could be due to shifts in values, but more likely it is due to turnover in election creating new governments. Italy between 2010 and 2014 shows this turnover.

Graph 4: GAL-TAN orientation for EU-countries, 1999-2019

Conclusion

One of the major findings in this project was that there is not as much movement among political parties in Sweden and EU over the years as one could think.In graph 1 that includes data from 1984, we could see general trends of ideological popularity. Conservative parties and Social democratic parties have lost voters to other parties like green, radical right and niche parties that have no real family belonging. Over time the political landscape has got more differentiated and diverse.

In graph 2 we could see a move to GAL for many Swedish parties, when faced by a growing radical right party. When the parties realize that they still looses voters they instead move closer to the radical right party in values to win back the voters. This effect could also be seen when looking at graph 3 indicating a overall movement for all parties in the analyzed countries. In graph 3 we could also see that the Swedish parties where more GAL oriented then their EU counterparts, mostly the agrarian party Centerpartiet. We also did not have any regional parties.

In graph 4 the movements are more diverse, but from 1999-2010 countries seemed to move more to GAL. After 2010 they seemed to move back to more TAN values. This finding correlates supports the evidence from graph 2 and 3 about how other parties react when faced by growing radical right parties.

Sources

Bakker, Ryan, Liesbet Hooghe, Seth Jolly, Gary Marks, Jonathan Polk, Jan Rovny, Marco Steenbergen, and Milada Anna Vachudova, (2020) “1999 − 2019 Chapel Hill Expert Survey Trend File.” Version 1.2. Available on chesdata.eu.

Leonard Ray (1999), “Measuring party orientations toward European integration: Results from an expert survey,” European Journal of Political Research, 36.2: 283-306.