[Blog] How close are the top British politicians to the electorate?

By Andrei Zhirnov, Michele Scotto di Vettimo & Susan Banducci (all University of Exeter).

How close are the top British politicians to the electorate?

Democratic elections are meant to ensure politicians promote the interests of voters – keeping them in office when they do and voting them out when they do not. Standard political science wisdom says that an astute politician will not advocate or make unpopular policies at the risk of being kicked out of office or otherwise jeopardizing their and their party’s political survival. However, the recent developments in British politics seem to suggest that it is not the case, and, at least according to the popular narrative, the current government has moved quite far from the public.

The data collected by a team of university researchers at Exeter, Münster, VU Amsterdam, Birmingham, and Rome – collectively known as PRECEDE – can help us assess how well the leading UK politicians represent the British electorate and the policy similarity we should expect in the near future.

In July 2022, around the resignation of Boris Johnson and the subsequent Conservative leadership contest, the PRECEDE consortium launched an online tool (Who Is Your Leader?) that allows internet users to answer a few questions about their attitudes to the political system and policies. In the end, they see their position in the political landscape along with those of the Conservative leadership contenders and other prominent contemporary and historical British political figures. Politicians and portal users are placed along two axes: the traditional left-right scale, which represents economic and cultural policies, and opposition to the political establishment. Politicians’ positions were estimated using a survey of experts in British politics.

Figure 1. Distribution of Users from the “Who Is Your Leader” Tool.



Figure 2. Distribution of Users that voted for the Conservative Party in 2019.


The tool also allows the researchers to take a bird’s eye view of the electoral landscape and see where most potential voters are sitting. Our early analyses of this data indicate a visible gap between the users of our tool and the Conservative leadership. Most of the users are further to the left than the Conservative party leaders. The distance is shorter for Rishi Sunak, but even he is quite far from the majority of the electorate.

If we focus only on the respondents who said they had voted for the Conservative Party in 2019, the distance becomes, as expected, smaller. Nevertheless, even in this case, the Labour leadership is almost as close to many Conservative voting users as the Conservative leadership.

This result is consistent with a YouGov study conducted around the same time as our data collection and a more recent one (October 20-21). Both report a lead for the Labour Party. In mid-August, 28% of decided voters intended to vote Conservative, and only 19% as of the 21st of October. The spatial voting approach, which assumes that the voters are more likely to vote for the parties closer to them spatially, seems to reconcile our observations about the policy landscape and vote intentions quite well.

However, something is puzzling about this current distance on the policy landscape: not long ago, in 2019, the Conservative party won a landslide, and it is hard to believe that it could be that far from the electorate at that time (or was it?). We have come up with a few possible explanations without giving any definitive answers.

  1. The pandemic and the cost of living crisis have left many people poorer, and even more, with a realization that the government needs to offer more support to those in economic distress. Consequently, voters have moved further to the left.
  2. The key issue of the 2019 campaign – the relationship between the UK and the EU – underpinned the attractiveness of the Conservative Party’s policies at the time but has lost its salience, and voters have drifted away.
  3. Voting is driven by affective support for a party label and specific candidates at least as much as voters’ policy attitudes. The parties occupied roughly the same quadrants of the policy landscape during the 2019 election as they do today. However, the respect for the “Conservative” label and the likeability of their leaders was much higher in 2019 than now.
  4. Tool users are systematically different from the electorate– they are more liberal. During our estimation procedure, we applied sampling weights to try and bring the distribution of user positions closer to the voting-eligible population. However, this cannot make our sample perfectly representative.

At this point, we have more questions than answers and hope that the outcomes of the new Conservative leadership contest and the next election will shed some light.



Italian Tool “Navigatore Populista” published

The PRECEDE Consortium has just published its tool for Italy, the Navigatore Populista: https://navigatorepopulista.it 

The tool consists of two dimensions (the axes of the political landscape). The horizontal axis summarises the statements that distinguish political “left” from political “right”, and the vertical axis differentiates between “populist” and “non-populist”. We have labeled the poles of the axes using keywords that are common when talking about politics. To position the parties on these axes, we administered a survey among academic experts. The survey included the same questions that we ask you in the tool. We aggregate these experts’ responses about political parties into their positions on our landscape in the same way as the users’ responses. Additionally, the tool also includes the positions of the party leaders, allowing users to compare parties and leaders – a unique feature that VAAs normally do not offer.

An overview of the party and leader positions can be found here: https://osf.io/z32sn/


British Party Leader Compass released

The PRECEDE Consortium is proud to present its latest Voting Advice Application, the Party Leader Tool “Who Is Your Leader”. The tool consists of two dimensions (the axes of the political landscape). The horizontal axis summarises the statements that distinguish political “left” from political “right”, and the vertical axis reflects the attitudes towards the roles of the elites and the public in government (pro and anti-establishment). We have labelled the poles of the axes using keywords that are common when talking about politics.

A team of experts from the University of Exeter was responsible for placing the party leaders on the statements. The survey included the same questions that we ask you in the tool. We aggregate these experts’ responses about political parties into their positions on our landscape the same way as the users’ responses.

The tool includes the positions of the two candidates in the final round of the leadership contest of the Conservative Party, Liz Tuss and Rishi Sunak. In addition, we also placed the most prominent contemporary and historical party leaders such as Margaret Thatcher, Tony Blair, Nigel Farage, Keir Starmer, Jeremy Corbyn and Boris Johnson.

The tool addresses the ongoing political debate in Great Britain and offers voters new insights in British politics.

[Blog] The Electoral Dynamics of the Hungarian Elections 2022

The Electoral Dynamics of the Hungarian Elections 2022

By Jan Philipp Thomeczek, Endre Borbáth, Alberto López Ortega, André Krouwel, Norbert Kersting, Kata Benedek


Introduction: Finally a Chance for the Opposition?

On Sunday, the most unpredictable Hungarian national election since 2006 will take place. After 12 consecutive years of Viktor Orbán and Fidesz majorities, the opposition has a realistic chance of approximating or even surpassing Fidesz’ vote share. Six opposition parties joined forces in the “United for Hungary” electoral coalition. During their primary elections, opposition voters elected a single candidate to face Fidesz in each of the 106 single member districts and united behind the prime minister candidate, Péter Márki-Zay. Mounting a centre-right challenge aimed to convince disenchanted Fidesz voters, Márki-Zay represents a centre-right position within a polyphonic big-tent opposition alliance. The six political parties of “United for Hungary” (UH) come from strongly different political camps: the radical right (Jobbik), centre(Momentum), centre-left (MSZP), left (DK), and ecologism (PM, LMP). As our analysis of the structure of party competition based on PRECEDE data shows, the parties of the opposition alliance are kept together by their strong anti-establishment, anti-Orbán stance.

The Orbán government prepared a referendum on legislation related to ‘child protection’ to confront the alliance in the campaign. The referendum is set to take place simultaneously with the election. The suggestive questions are widely considered as homophobic, as it seems impossible to reject them. This plebicscite seems to lack integrity in different areas, if one would apply the criterium of other referendums. Given the overlap between the referendum and the elections, gender issues became a key issue of the campaign. However, this suddenly changed after the Russian invasion of Ukraine. The war has forced the Orbán government to change its strategy of fighting the campaign on gender. 

After 12 years of investing in a close alliance with Russia, the policies of the Fidesz government came under scrutiny after Russia invaded Ukraine. Those policies include extending the Paks nuclear power plant with Russian loans, hosting the Russian backed International Investment Bank (often seen as a Trojan horse of Russian intelligence operations in Europe), regular Putin-Orbán meetings, and the increasingly close business cooperation with the Russian elite. With NATO and EU alliances closing ranks in solidarity with Ukraine, Orbán partly reversed his government’s uncritical stance on Russia and identified a pro-peace, pro-neutrality position as the electorally least harmful alternative. Meanwhile, the “United for Hungary” alliance consistently represents a political position more closely aligned with the NATO and EU countries against Russia. In the last month of the campaign, the war became a “super-issue”, crowding out the coverage of other issues in the domestic press.


The Positions of the Hungarian Political Parties and Coalitions

Party Expert Position Survey (PEPS) Hungary

Where are the most important Hungarian parties and the new oppositional coalition “United for Hungary” located ideologically? To position the political parties, the PRECEDE project conducted an expert survey among Hungarian researchers who specialised in Hungarian politics. Since experts had to select the parties they wanted to place, the number of responses per party and issue differs, with around 30 placements for the biggest parties. The party positions were identified using the median of the expert placements, as it is less strongly influenced by outliers than the mean. Each of the statements was then assigned to the dimensions of interest. To identify the party positions per dimension, the mean of all statements assigned to the dimension was calculated, as frequently used by Voting Advice Applications based on a two-dimensional space, including other PRECEDE tools. We recently developed several Voting Advice Applications for other countries, including the Netherlands, Germany, and France, using a similar method. 


Cultural and Economic Issues

Figure 1 shows the distribution of party positions based on the two most salient dimensions of European party politics, an economic left-right dimension and a cultural/societal GAL-TAN (Green Alternative Libertarian – Traditional Authoritarian Nationalist) dimension. As the dimension on European Integration aligns well with the GAL-TAN dimension, we decided to combine the two dimensions for simplicity, as other Voting Advice Applications such as EU and I have previously done before.

Two aspects of the Hungarian political landscape become apparent by looking at Figure 1. First, the political competition on the economic dimension is relatively centrist. All oppositional groups are somewhat centre-left, thus in moderate favour of a “strong state”. Momentum, part of the liberal ALDE European party association, takes the least leftist position within the UH coalition. Fidesz-KDNP takes the most market-friendly positions, with still relatively moderate positions. Overall, the economic dimension is structured by many valence issues, where parties are rather concerned with the way and intensity an issue needs to be solved instead of its general direction. This is also reflected in economic statements from the pre-selection where all parties were “on the same side” (e.g. increasing the minimum wage or increasing the wage of civil servants). Nevertheless, there may be a discrepancy between the expert assessment and official party communication for some issues.

The second main conclusion drawn from Figure 1 is that political competition is strongly polarised along the GAL-TAN dimension. While all UH coalition parties except for Jobbik, plus satire party MKKP, are located in the GAL quadrant, Fidesz and right-wing populist newcomer Our Homeland are clearly in favour of TAN positions. Jobbik is located somehow in between, quite distant from its other UH coalition partners. By looking at Figure 1, one may wonder why ideological clearly different parties such as Momentum and Jobbik would officially team up in an electoral coalition. The answer to this question can probably be found in Figure 2. 

Figure 1. Political Landscape of Hungary based on an Economic and a GAL-TAN Dimension.


Pro vs Anti-Establishment

In Figure 2, the vertical axis is the same as in Figure 1. The horizontal dimension now summarises statements dividing parties into a Pro- and Anti-Establishment dimension. Those statements relate to topics such as anti-elitism, and national-specific political conflicts about the government’s concentration of power (e.g. influence on the selection of the cultural and scientific elite, independence of state media, anti-corruption measures and autonomy of municipalities). The positions of the political parties on this dimension show clearly that the primary purpose of the United for Hungary alliance is to defeat Fidesz. All UH parties are united in their anti-government stances, including Jobbik. The non-UH opposition, in contrast, is divided. While MKKP is strongly anti-government, Our Homeland takes more nuanced positions. Nevertheless, even Our Homeland is located on the anti-government side.

Figure 2. Political Landscape of Hungary based on an Anti-/Pro-Establishment Dimension and a GAL-TAN Dimension.


European Integration

During the last decade, researchers have observed growing democratic backsliding and the transformation towards an illiberal democracy in Hungary, reflected the many conflicts that the Hungarian government, led by Viktor Orbán, provoked with the European Union. From a Western European perspective, one may think that this stems from or is at least correlated with growing Euroscepticism in Hungary. However, looking at the party positions towards EU membership, this does not seem to be the case. The experts classify only one party, namely far-right Our Homeland, as clearly anti-EU; all other oppositional parties are not Eurosceptic. To external observers, it may be surprising that Fidesz is classified as “neutral” towards the statement that EU membership has “mostly been bad for Hungary”. However, up to date, a potential “Huxit” was officially never suggested by Fidesz. Orbán’s fight with the European Union is primarily rhetorical. A recent poll of 2020 showed that leaving the European Union is only supported by a small minority (17%) of the Hungarian population, suggesting that even Fidesz voters are against this measure. Instead, Western leaders have occasionally suggested that Hungary may be better off leaving the EU. For example, in 2021, after the Fidesz government had passed a very controversial Anti-LGBTQI law, Dutch prime minister Rutte suggested that Hungary may consider leaving the EU if it did not respect EU values.


Party/Coalition Overall, EU membership has been a bad thing for Hungary.
Fidesz Neutral
EM Coalition Strongly Disagree
Jobbik Disagree
MSZP Disagree
PM Disagree
DK Disagree
LMP Strongly Disagree
MM Disagree
Our Homeland Strongly Agree
MKKP Disagree

Table 1. Party Positions toward European Integration. 



While the election outcome remains uncertain, Fidesz is the clear favourite. In recent years, Orbán has successfully shaped an electorate homogenous in its preferences in a highly polarised space, where a defection of Fidesz voters to the opposition parties remains the exception. The opposition took the first hurdle with success by joining forces, giving them a real chance at winning for the first time since 2006. The UH coalition is primarily held together by their strong anti-Orbánism. Nevertheless, the six-party coalition is fragmented both organisationally and programmatically. To successfully face the institutional limits that stay in place even if they win the election and form a stable government, they need to strengthen both their ideological and organisational cohesion. In the last days, a poorly phrased sentence, a picture from the past of a Jobbik politician, or a reminder of the controversial role previous left-wing prime minister Gyurcsány played in the 2006 protest are still able to dissuade voters of opposition parties.






Statement Scaling
Quite a few of the people running the government are crooked. Anti-Establishment
The government is pretty much run by a few big interests looking out for themselves. Anti-Establishment
MPs in Parliament very quickly lose touch with ordinary people. Anti-Establishment
Hungary should set up an Anti-Corruption  Commission. Anti-Establishment
There is too much political influence on the selection of cultural and scientific directors. Anti-Establishment
The state media should be made more independent from political influence. Anti-Establishment
Municipal authorities need more autonomy. Anti-Establishment
Hungary needs a new Constitution. Anti-Establishment
Government officials use their power to try to improve people’s lives. Pro-Establishment


The state should extend the financial support of unemployed workers from 3 to 9 months. Economic Left
The tax rate for the highest incomes must be increased. Economic Left
Wealth should be redistributed from the richest people to the poorest. Economic Left
In general, government spending should be reduced in order to lower taxes. Economic Right
In general, the state should intervene as little as possible in the economy. Economic Right
In order to keep pensions affordable, the retirement age may be raised if life expectancy rises. Economic Right
Workers’ protection regulations should be reduced to fight unemployment. Economic Right
It should be easier to have an abortion. GAL
A party’s seat share should be fully determined on the basis of its vote share (full proportional representation). GAL
The people I disagree with politically are not evil. GAL
Renewable sources of energy should be supported even if this means higher energy costs. GAL
Churches should receive less public funding. GAL
Major government decisions should be approved by a referendum. GAL
Stricter climate legislation must be introduced, even if it stands in the way of economic growth. GAL
It should be compulsory for children to go to school until the age of 18. GAL
Only citizens residing in Hungary permanently should vote in the Hungarian elections. GAL
Women are disadvantaged in our society. GAL
Members of the LGBT community threaten the institutions of the traditional family. TAN
You can tell if a person is good or bad if you know their politics. TAN
Immigration to Hungary should be made more restrictive. TAN
The climate crisis is exaggerated. TAN
Immigrants should adapt to Hungarian values. TAN
Only legally recognized traditional families (married heterosexual couples) should be helped financially by the state. TAN
Criminals should be punished more severely. TAN
It is not possible to integrate the Roma population into Hungarian society. TAN
Sex education is the responsibility of parents, not of schools. TAN
Membership of the EU has so far been mostly bad for Hungary. TAN
Christian values should be given a more central place in political decision-making. TAN
The people I disagree with politically are just misinformed. TAN
The adoption of the single European currency (EURO) would be a bad thing for Hungary. TAN

Table 2. List of Statements and Scaling. Statements in Italic are not included in the Tool.

[Blog] The 2022 Hungarian Elections in the Shadow of the Russian War in Ukraine

Endre Borbáth, Jan Philipp Thomeczek, Alberto López Ortega, Norbert Kersting, André Krouwel

 (For the German version of this article, please click here).

On the upcoming Sunday, the most unpredictable Hungarian national election will take place since 2006. After 12 consecutive years of Viktor Orbán and Fidesz majorities, the opposition has a realistic chance of approximating or even surpassing Fidesz’ vote share. Following a civil society organized primary election, six opposition parties came together and joined forces. During the primary elections, opposition voters elected a single candidate to face Fidesz in each of the 106 individual member districts and united behind the prime minister candidate, Péter Márki-Zay. Mounting a centre-right challenge aimed to convince disenchanted ex-Fidesz voters, Márki-Zay represents a centre-right position within a polyphonic big-tent opposition alliance. The six political parties of the alliance ‘United for Hungary’ previously occupied radical right positions (Jobbik), centrist positions (Momentum), centre-left positions (MSZP), left positions (DK), as well as pro-environmental positions (PM, LMP). Beyond the pressure of the majoritarian electoral system, the parties of the opposition alliance are bonded together by their strong anti-establishment, anti-Orbán stance.


The campaign was shaping up to be a ‘regular’ election campaign, dominated by ‘socio-cultural’ issues such as the politics of gender. The government called for a referendum to be held at the same time as the election on legislation related to ‘child protection. This plebiscite seems to lack integrity in different areas, if one would apply the criterium of other referendums.  The questions are highly suggestive, widely considered homophobic designed to back opposition parties into a corner. However, the dynamic was suddenly changed by the launch of the Russian attack on Ukraine. The war has forced the Orbán government to change its strategy of fighting the campaign on gender. After 12 years of having invested in a close alliance with Russia, the Fidesz government’s policies came under close scrutiny. This includes the agreement to extend the Paks nuclear power plant based on a Russian backed loan, hosting the Russian backed International Investment Bank (often seen as a Trojan horse of Russian intelligence operations in Europe), hosting regular Putin-Orbán meetings, and investing in close business corporations with Russian elites. With the NATO and EU alliances closing ranks in solidarity with Ukraine, Orbán partly reversed his government’s uncritical stance on Russia and found a pro-peace, pro-neutrality position, as the electorally least harmful alternative. As part of this strategy, the government remains reluctant to impose hard sanctions on Russia or to support military efforts, such as the transport of weapons undertaken by the NATO alliance, but it also shies away from openly challenging NATO. Some observers suggest that Hungary may not even allow NATO planes to cross their territory. Meanwhile, the ‘United for Hungary’ alliance consistently represents a political position more closely aligned with the NATO and EU countries, against Russian aggression. In the last month of the campaign, the war became a ‘super-issue’ crowding out the coverage of other domestic political concerns in the press.


Although new in its current manifestation, parties’ stance on the war feeds into the broader dimension of pro- and anti-EU positions. Based on a survey of 80 experts, the PRECEDE project’s voting advice application shows Fidesz’s neutral position on the question of overall evaluating EU membership of Hungary, for example. Despite high popular support for EU membership, since in government Fidesz has taken an ambivalent position on the EU. While it is constantly engaged in a ‘fight against Brussels’, Orbán previously argued that Hungary is entitled to the benefits of EU membership as an emerging market that provides investment opportunities for Western European economies. Hungarian exit has never been on Fidesz’ agenda, instead, he argues for ‘reforming’ the Union to more closely reflect what he considers to be the ‘European’ heritage. The opposition alliance takes a less critical position, with prime minister candidate Márki-Zay promising to introduce the Euro soon. Under the conditions of the current war, Orbán takes a similar ‘pragmatic’ position, arguing that ‘Hungarian interests’ trump more principled views on Ukrainian solidarity. His position remains tenable to the extent that no Russian attack is imminent on a NATO-allied country. Opposition parties frame the country’s role in the war as a test of Hungary’s commitment to the ‘Western world’.


While the outcome of the election remains uncertain, Fidesz is a clear favourite. The few public opinion polls that have been conducted since the outbreak of the war show that many voters agree with the pro-peace position Viktor Orbán has taken. On the national day on the 15th of March, activists diffused a previous speech by Viktor Orbán to the marching pro-government supporters in which he argued for a principled view against an autocrat such as Vladimir Putin. Government supporters booed this prior speech, displaying a remarkable commitment to the prime minister that goes beyond a demand for consistent position-taking. Orbán has successfully shaped an electorate homogenous in its preferences in a highly polarized space, where Fidesz-voter defection to the opposition parties remains exceptional. Should the opposition alliance emerge as victorious, to successfully face the institutional limits that stay in place even if they win the election, and form a stable government for years to come, they need to make similar investments in both ideological and organizational cohesion. At present, a poorly phrased sentence, a picture from the past of a Jobbik politician, or a reminder of the role previous left-wing prime minister Gyurcsány played in crushing the 2006 protest are still able to dissuade voters of opposition parties.

Hungarian Election Compass “Választási Iránytű” released

The PRECEDE Consortium is proud to announce the release of our Voting Advice Application “Választási Iránytű”, designed for the Hungarian Elections on April 3rd. The tool can be accessed via  http://valasztasi-iranytu.org/.

In order to determine the position of the parties towards the different issues, a survey among scientific more than 80 Hungarian party experts was used as a source. The survey is based on the questions from the 2022 Party Expert Positioning Survey (PEPS). The same questions are included in the tool. The party positions can be downloaded here.

The Election Compass Hungary offers voters unique information in a peculiar electoral environment.  The tool includes both the electoral coalitions as well as the positions of the individual parties. Thus, voters can use the tool to identify how close the individual parties that make up the electoral coalitions are.

French Tool “Vox Populi” released

In collaboration with our colleagues Thomas Vitiello and Bruno Cautrès (Scienes Po Paris) we have recently released the tool “Vox Populi” for France. After answering 23 statements, respondents can explore their position in a political landscape consisting of classic Left-Right as well a Populism dimension. In addition, the most important presidential candidates have been placed on the same statements, allowing respondents also to see where the different candidates are located politically. The Vox Populi tool, therefore, enables voters to see where the candidates stand relatively early in the presidential election campaign.

The tool can be accessed via https://voxpopuliste.fr/ (in French)

German Tool Polit-Kompass released

We are happy to announce that our latest tool, the German Polit-Kompass, has just been released. With the Polit-Kompass, we have prepared the first out of two tools we are developing for Germany. The second tool, the Wahl-Kompass, will be released in August 2021.

The Polit-Kompass is the sister tool of the Dutch Populismemeter.  After answering 30 statements by using a five-point agreement scale, users will get an individual position in the political landscape, where also the most relevant German parties are positioned.

For the German users, we summarised background information here as well as the scientific background behind the tool in this blog article.

[Blog] Der Polit-Kompass: Eine neue VAA zum aufkeimenden Populismus [German]

Erschienen auf Regierungsforschung.de (CC BY-NC-SA Lizenz)

Von Jan Philipp Thomeczek, Norbert Kersting, André Krouwel, Susan Banducci, Lorenza Antonucci, Carlo D’Ippoliti, Andrei Zhirnov, Laszlo Horvath, Christian Mongeau

Mit der Covid19-Pandemie haben bei Wahlen digitale Beteiligungsinstrumente und insbesondere Onlinewahlhilfen (Voting Advice Application, VAA) einen besonderen Stellenwert erlangt (Kersting 2021). Mit dem Polit-Kompass wurde ein Tool entwickelt, das bei der politischen Orientierung unterstützen soll und gleichzeitig eng an die politikwissenschaftliche Forschung angebunden ist. Dabei umfasst er neben einer allgemeinen Links-Rechts-Dimension auch eine Populismus-Dimension

Der Polit-Kompass: Eine neue VAA zum aufkeimenden Populismus

Mit der Covid19-Pandemie haben bei Wahlen digitale Beteiligungsinstrumente und insbesondere Onlinewahlhilfen (Voting Advice Application, VAA) einen besonderen Stellenwert erlangt (Kersting 2021). Mit dem Polit-Kompass wurde ein Tool entwickelt, das bei der politischen Orientierung unterstützen soll und gleichzeitig eng an die politikwissenschaftliche Forschung angebunden ist. Dabei umfasst er neben einer allgemeinen Links-Rechts-Dimension auch eine Populismus-Dimension, die in den letzten Jahren stark an Bedeutung gewonnen hat. Der Polit-Kompass soll dabei allgemeine politische Konflikte abdecken, die zu Wahlkampfzeiten besondere Brisanz erfahren, aber eben auch außerhalb von Wahlen für Diskussionen sorgen.


Die Kompass-Familie


Online-Wahlhilfe Tools wie der Polit-Kompass werden in der politikwissenschaftlichen Forschung als Voting Advice Applications (VAAs) bezeichnet (Thomeczek, im Erscheinen). Als „issue matching systems“ (Ladner & Fivaz, 2012, S. 178) vergleichen sie die Positionen der Nutzerinnen und Nutzer mit den Positionen des politischen Angebots, in diesem Fall also von Parteien. Das Ergebnis basiert dabei auf der Zustimmung zu verschiedenen, meist zwischen 20 und 40, gerichtet formulierten Thesen, für die auch Parteipositionen vorliegen. Die Ergebnisse werden zunächst für jede These einzeln berechnet. In mehrdimensionalen Räumen kann das Ergebnis auch direkt als Distanz zu den einzelnen Parteien dargestellt werden.

Bei der Aggregierung der Ergebnisse wählen VAAs unterschiedliche Ansätze, die als High Dimensional Matching und Low Dimensional Matching bezeichnet werden (Mendez, 2014). Im High Dimensional Matching stellt jede These eine eigene Dimension dar. Als Gesamtergebnis wird hier meist ein Ranking mit einem Prozentwert dargestellt, der angibt, wie stark die Übereinstimmung gemessen an der maximal möglichen Übereinstimmung ausfällt. In Deutschland nutzen beispielweise der Wahl-O-Mat der Bundeszentrale für politische Bildung oder der WahlSwiper diese Darstellungsmethode. Werden die Positionen zu einer oder mehreren Dimensionen aggregiert, spricht man vom Low Dimensional Matching. Hier werden die Thesen den verschiedenen Dimensionen zugeordnet, sodass für jede Dimension eine eigene Position ermittelt werden kann. Besonders weit verbreitet ist hierbei eine zweidimensionale Darstellung, die beispielsweise von der Kieskompas-Familie (Krouwel et al., 2012), zu der auch der deutsche Wahl-Kompass gehört (Thomeczek et al., 2019), oder dem schweizerischen Smartvote genutzt wird. Ein zweidimensionales Koordinatensystem erlaubt dabei eine enge Anbindung an die politikwissenschaftliche Forschung (Marks et al., 2006).


Auswahl und Zuordnung der Thesen


Im PRECEDE Projekt steht der länderübergreifende Vergleich im Vordergrund. Um die Positionen von Parteien und Wahlberechtigten in zehn Ländern (Niederlande, Rumänien, Deutschland, Ungarn, Frankreich, Tschechien, Spanien, Großbritannien, Italien und Schweden) miteinander zu vergleichen, mussten wir zunächst Thesen identifizieren, die für einen solchen Vergleich geeignet sind und die politischen Konflikte in allen Ländern gut abbilden können. Aus diesem Grund wurden zunächst Daten der europäischen VAA EU and I analysiert. Das Projekt bietet bereits seit 2009 (damals unter dem Namen EU Profiler) VAAs zur Europawahl an (Reiljan et al., 2020). Die Positionsdaten der Parteien werden für alle Mitgliedstaaten erhoben, sodass 2019 unsere zehn Länder abgedeckt wurden. Mithilfe dieser Daten können also länderübergreifende politische Dimensionen identifiziert werden.

Mithilfe von dimensionsreduzierenden Verfahren wurden zunächst besonders hochskalierende Thesen auf der Ebene der Partei und der Ebene der Nutzerinnen und Nutzer identifiziert. Auf Basis der EU and I-Daten zeigt sich, dass eine zweidimensionale Verortung einen Großteil der Information sinnvoll zusammenfassen kann. Diese beiden Dimensionen lassen sich aufgrund theoretischer Überlegungen als Pro/Anti-EU-Dimension und Links/Rechts-Dimension bezeichnen. Erstere stellt dabei eine Besonderheit der VAA dar, da viele EU-spezifische Thesen vorliegen. Die Links-Rechts-Dimension umfasst dabei sowohl wirtschaftliche als auch gesellschaftliche Positionen, ist also eine allgemeine „Super Dimension“,  auf der sich die meisten Konflikte einordnen lassen und die in der Forschung immer wieder diskutiert wurde (Gabel & Huber, 2000). Dies steht im Kontrast zu den meisten EU VAAs wie EU and I und die VAAs der Kieskompas-Familie zur Europawahl, die die EU-Dimension in die gesellschaftliche Dimension integriert und die Links-Rechts-Dimension auf den wirtschaftlichen Konflikt beschränkt hatten.

Da das PRECEDE-Projekt in besonderer Weise auf die Messung von populistischen Positionen abzielt, wurden zudem Populismus-Items aus der Individualforschung aufgenommen (Castanho Silva et al., 2020). Hierbei wurde sich am ideellen Ansatz der Populismusforschung orientiert (Mudde, 2017), der den Konflikt zwischen zwei homogenen Gruppen zwischen Establishment/Elite und dem Volk/der Bevölkerung in den Vordergrund stellt. Die Wichtigkeit dieser neuen populistischen Konfliktlinie hat in den letzten Jahren stark zugenommen (Norris & Inglehart, 2019; Schäfer & Zürn, 2021). Dies trifft insbesondere auch auf die politische Landschaft in Deutschland zu, die seit 2013 durch eine populistisch auftretende AfD beeinflusst wurde.

Der so gebildete Thesen-Pool wurde mit Thesen aus weiteren Analysen und Umfragen zu neueren Entwicklungen (beispielsweise im Bereich der Umweltpolitik) ergänzt. Ziel ist es dabei, in den jeweiligen Tools der zehn Länder eine möglichst hohe Überlappung an Thesen zu erreichen. Gleichzeitig erkennen wir dennoch an, dass einige politische Konflikte zu nationalspezifisch sind, wofür individuelle Thesen nötig sind. Die wahlspezifischeren Themen werden durch das zweite Tool, welches wir zur Bundestagswahl entwickeln, den Wahl-Kompass, stärker in den Fokus genommen (Veröffentlichung im August 2021).


Expert Survey PEPS


Zur Bestimmung der Parteipositionen im Polit-Kompass wurde eine Befragung unter Expertinnen und Experten (Party Positioning Expert Survey, PEPS) durchgeführt. Hierzu wurden 237 Personen, die zu Parteien in Deutschland forschen, zur Befragung eingeladen. 87 Personen haben die Befragung mindestens begonnen, wobei die Befragten auswählen konnten, welche Parteien sie positionieren möchten und Thesen überspringen konnten. Zur Auswahl standen alle Parteien, die mindestens einen Sitz bei der Europawahl 2019 gewonnen hatten – neben den Bundestagsparteien also auch Die PARTEI, Piraten, Volt, Freie Wähler, ÖDP, Familien-Partei und Tierschutzpartei. Die Befragung hatte zum Ziel, die verschiedenen Parteien bezüglich der ausgewählten Thesen zu positionieren, also die Thesen aus Sicht der Partei zu beantworten, sodass diese zur Berechnung der Parteiposition im Polit-Kompass zugrunde gelegt werden kann. Hierzu stand eine fünfstufige Likert-Skala zur Verfügung. Auf der (inhaltlichen) Policy-Ebene umfassten die Thesen Positionen der wirtschafts- und sozialpolitischen Dimension, der gesellschaftlich-kulturellen Dimensionen und der EU-Dimension. Darüber hinaus haben wir auch einige Survey-Items der Populismusforschung auf Individualebene für unsere Umfrage ausgewählt. Insbesondere die Thesen zur Populismus-Dimension waren in der Befragung für uns von großem Interesse, da diese mit einer nachweisbasierten Kodiermethode der Parteipositionen, die wir für den Wahl-Kompass normalerweise anwenden, mangels konkreter Textnachweise nicht umzusetzen wäre; hierzu sind die Populismus-Items aus der Survey Forschung in der Regel zu spezifisch formuliert (zum Beispiel: „Regierungsmitglieder nutzen ihre Macht, um das Leben der Menschen zu verbessern.“). Diese Methode wurde von PRECEDE erstmals für die Positionierung von Parteien angewendet, wobei ein ähnliches Vorgehen bereits für ein Guardian-Projekt mit der Positionierung von internationalen Politikerinnen und Politikern umgesetzt wurde (Lewis et al., 2018). Gleichzeitig können so die Positionierungen der Parteien mit den Positionen der (potenziellen) Wählerinnen und Wählern verglichen werden.


Generierung der Parteipositionen im Polit-Kompass


Die individuellen Positionen der Befragten je Partei und These mussten zunächst aggregiert werden. Andere politikwissenschaftliche Expertenumfragen, wie die regelmäßig durchgeführte Chapel Hill Expert Survey (Bakker et al., 2019), nutzen dazu häufig den Mittelwert der Positionen. Da der Mittelwert jedoch anfällig für Ausreißerwerte ist, sind andere Maße der zentralen Tendenz wie der Modalwert oder der Median zu bevorzugen (Lindstädt et al., 2020). Allerdings hat dies bei reinen Ordinalskalen den Nachteil, dass möglicherweise keine eindeutigen Werte bestimmt werden können, da mehrere Werte in Frage kommen können. Beim Polit-Kompass wurde daher ein mehrstufiges, hierarchisch aufgebautes Verfahren zur Aggregierung der Parteipositionierungen eingesetzt. Zunächst wurden alle Thesen-Partei-Kombinationen von der Aggregierung ausgeschlossen, für die weniger als fünf Positionierungen vorlagen. Die umfasst die Tierschutzpartei und die Familien-Partei vollständig, die daher nicht aufgenommen wurden, sowie einige Thesen für die kleineren Parteien, die nicht im Bundestag vertreten sind. Im ersten Schritt haben wir die mehrheitlich ausgewählte Antwortkategorie (Modalwert) der Expertinnen und Experten als Parteiposition herangezogen. In den Fällen, in denen dies nicht möglich war, weil mehrere Antwortkategorien gleichhäufig ausgewählt wurden, wurde der Median berechnet. Wenn auch der Median nicht eindeutig bestimmt werden konnte, weil er genau zwischen zwei Antwortkategorien lag, haben wir den Mittelwert herangezogen. Mithilfe des Mittelwerts haben wir in dem Fall eine Rundungstendenz ermittelt (Auf- oder Abrundung), sodass in jedem Fall eine der fünf Antwortkategorien als finale Parteiposition bestimmt werden konnten.

Um aus den Thesenpositionen der Parteien eine politische Landschaft zu entwickeln, ist eine Zuordnung der einzelnen Thesen zu den zwei Dimensionen notwendig. Aufgrund der Analysen der EU and I-Daten und inhaltlicher Überlegungen lag bereits eine A-Priori-Verortung vor. Da nun jedoch die Populismus-Thesen neu aufgenommen wurden und eine Reduzierung auf 30 Thesen angestrebt wurde, wurde die Zuordnung erneut empirisch überprüft. Hierbei hat sich gezeigt, dass eine zweidimensionale Verortung einen Großenteil der Informationen zusammenfassen kann. Die Links-Rechts-Dimension fungiert hier wieder als „Super Dimension“, die sowohl wirtschafts-sozialpolitische als auch gesellschaftlich-kulturelle Konflikte umfasst. Auf der zweiten Dimension sind die Populismus-Thesen und EU-Thesen verortet. Hier zeigt sich, dass EU-kritische Haltungen mit populistischen Einstellungen einhergehen.

Dass Populismus- und EU-bezogenen Thesen auf einer Dimension verortet sind, mag zunächst überraschen, da die meisten Populismus-Definitionen wie der ideelle Ansatz keine EU-Dimension einschließen. Viele populistische Parteien stellen die EU jedoch regelmäßig als elitäres Konstrukt dar. Der Zusammenhang wird auch bei einer Analyse in Abbildung 1 sichtbar. Hier sind zwei Variablen, der Populismusgrad und die Unterstützung für die Europäische Integration, aus dem POPPA, dem Populism and Political Parties Expert Survey (Meijers & Zaslove, 2020), dargestellt, der weiter unten auch zur Validierung unserer Ergebnisse herangezogen wird. Obwohl auch sowohl EU-freundliche populistische Parteien als auch EU-kritische unpopulistische Parteien existieren, wird deutlich, dass die meisten populistischen Parteien sehr negativ gegenüber der europäischen Integration eingestellt sind.

Abbildung 1. Zusammenhang zwischen Populismusgrad und Unterstützung für die Europäische Integration im Populism and Political Parties Expert Survey (POPPA).


Die politische Landschaft im Polit-Kompass


Die politische Landschaft in unserem Polit-Kompass besteht folglich aus zwei Dimensionen: einer Links-Rechts „Super Dimension“, die die wichtigsten politischen Fragen in rechte und linke Positionen einteilt, und eine Populismus-EU Dimension, die zwischen populistisch/EU-kritisch und unpopulistisch/EU-freundlich unterscheidet. Zur Berechnung der Positionen nutzt der Polit-Kompass den Mittelwert, wozu umgekehrt formuliert Thesen rekodiert wurden. Durch die Kombination der beiden Positionen ergibt sich die politische Landschaft, die in Abbildung 2 dargestellt ist.

Im Quadranten rechts unten, der unpopulistische und rechte Positionen kombiniert, ist das „bürgerliche Lager“ mit CDU, CSU und FDP positioniert. Von den drei Parteien ist die CSU am weitesten rechts positioniert, exakt zwischen dem rechten Pol und der Mitte. Die FDP ist näher an der Mitte verortet als die Unionsparteien. Sie vertritt klar wirtschaftsliberale Positionen, beispielsweise in Bezug auf den Kündigungsschutz, Steuererhöhungen beziehungsweise -senkungen und Umverteilung. Allerdings wird sie durch gesellschaftspolitisch liberale Positionen, die CDU und CSU meist nicht vertreten, „nach links“ gezogen. Beispielhaft für gesellschaftspolitische liberale Positionen stehen die Positionen der FDP bei den Themen Abtreibung, Cannabislegalisierung und Adoption durch gleichgeschlechtliche Paare. Die Freien Wähler nehmen ähnliche Mitte-Rechts Positionen wie die CDU ein, sind aber auf der vertikalen Populismus-Achse deutlich weiter oben, in etwa auf Höhe der Mitte, positioniert. Im Vergleich zu CDU, CSU und FDP sind sie also zweifelsohne populistischer, aber keineswegs klar populistisch verortet. Die AfD steht auf der Links-Rechts-Achse noch weiter rechts als die vier Parteien, setzt sich aber vor allem dadurch ab, dass sie stark populistische positioniert ist. Sie stimmt fast allen Populismus-Thesen in vollem Umfang zu beziehungsweise lehnt umgekehrt formulierte Thesen stark ab. Auch wenn verschiedentlich die inhaltliche Nähe von AfD und Unionsparteien in einigen Punkten betont werden, so zeigt sich hier deutlich, wie weit diese auf der Populismus-Dimension auseinander liegen.

Abbildung 2. Zweidimensionale politische Landschaft im Polit-Kompass.

Im „linken Lager“ ist das Bild etwas diffuser. Zunächst ist auffällig, dass mit Ausnahme der Freien Wähler auch alle Kleinstparteien (Piraten, Die PARTEI, ÖDP und Volt) links der Mitte verortet sind. Grüne, Die Partei und Die Linke stehen auf der Links-Rechts-Achse am weitesten links. Dies bedeutet jedoch nicht, dass beide Parteien identische Positionen vertreten. Die Grünen sind im Bereich der wirtschafts- und sozialpolitischen Thesen etwas gemäßigter positioniert, wohingegen dies auf Die Linke bezüglich der Themen Umweltschutz und Migration zutrifft. Letzteres ist sicherlich auch Ausdruck des Umstandes, dass die ehemalige Fraktionsvorsitzende Wagenknecht die innerparteiliche Debatte hierzu immer wieder angeheizt hat. Volt und Piraten sind insgesamt näher an der Mitte verortet als Grüne und Linke, SPD und ÖDP noch weiter. Auf der vertikalen Populismus-Achse zeichnen sich zwei Pole ab. Die SPD wird hier als am wenigsten populistisch verortet und nimmt eine ähnliche Position auf der Dimension wie die CSU ein. Am anderen Ende steht Die Linke, deren Position in etwa zwischen dem populistischen Extrempol und der Mitte liegt. Damit ist sie innerhalb des linken Lagers die populistischste Partei, jedoch deutlich weniger populistisch als die AfD verortet. Die Grünen und Volt sind unterhalb der Mittellinie positioniert, Die PARTEI, Piraten und ÖDP etwas über dieser Grenze. Sie vertreten hier gemischte Positionen, die sich gegenseitig ausgleichen.


Validierung der Parteipositionen im Polit-Kompass


Inwiefern kommen andere politikwissenschaftliche Expert Surveys zu ähnlichen Ergebnissen? Hierzu wurden die Einordnungen im Polit-Kompass mit den Einordnungen aus drei anderen Projekten, der PopuList (Rooduijn et al., 2020), der Populist and Political Parties Expert Survey (POPPA, Meijers & Zaslove, 2020) und der Chapel Hill Expert Survey (CHES, Bakker et al., 2019). Bevor die Ergebnisse dieser Projekte miteinander verglichen werden, müssen die Unterschiede der Befragungsformen diskutiert werden. Die Positionierung von Parteien anhand spezifischer Statements wird zum Teil im Kontext von VAA-Entwicklungen angewendet (zum Beispiel im Rahmen der sogenannten Delphi-Methode, siehe Gemenis, 2015), ist für Befragungen unter Expertinnen und Experten aber eher unüblich. Zur Entwicklung von VAAs ist sie jedoch nötig, um die Übereinstimmung zwischen Parteien und Nutzerinnen und Nutzern zu berechnen. Die meisten Befragungen fragen hierzu die Einstufungen auf den verschiedenen Dimensionen direkt ab, in der Regel mit elfstufigen Skalen. CHES und POPPA erheben zusätzlich zu einer allgemeinen Links-Rechts Dimension, wie sie auch im Polit-Kompass genutzt wird, eine wirtschaftliche Achse; CHES zudem auch eine eigene gesellschaftliche Dimension (GAL-TAN). Der CHES erhebt zwar weitere Subdimensionen zu einzelnen Policies wie zur staatlichen Intervention, zum Umweltschutz oder zur Unterstützung von Multikulturalismus, jedoch werden diese unabhängig von der Gesamteinstufung erhoben. Der Polit-Kompass hingegen aggregiert die Einzelstatements zu einer Dimension. Hierbei fließen keine Gewichte ein – anders als Beispiel beim Populismusgrad in POPPA, der sich aus der gewichteten Einordnung der Subdimensionen „Manichean“, „Indivisble“, „Generalwill“, „Peoplecentrism“ und „Antielitism“ zusammensetzt (Meijers & Zaslove, 2020). Zur Vergleichbarkeit wurden alle Variablen reskaliert, so dass diese nur Werte zwischen 0 und 100 einnehmen können.

In Abbildung 3 ist der Vergleich der Links-Rechts Einordnung im Polit-Kompass (Y-Achse) mit der allgemeinen Links-Rechts Einordnung (lrgen) im CHES dargestellt. Die gestrichelten Linien zeigen jeweils den mittleren Skalenpunkt an, die schwarze Diagonale den perfekten Zusammenhang, der vorliegt, wenn die (reskalierte) Position aus dem Polit-Kompass exakt der Position im CHES entspricht. Es ist zu erkennen, dass die Parteipositionen relativ nahe in der Diagonalen liegen. Die größten Abweichungen liegen für die Grünen vor, die im Polit-Kompass linker eingestuft werden. Dieser Umstand könnte auf zwei Faktoren zurückzuführen sein. Zum einen nehmen umweltpolitische Statements (2/18) aufgrund der steigenden Bedeutung dieses Themas einen relativ großen Raum im Polit-Kompass ein. Zum anderen könnte dies auch eine tatsächliche Positionsverschiebung der Grünen andeuten. Nichtsdestotrotz sind die Abweichungen insgesamt gering, selbst die Kleinstpartei Piraten, für die weniger Einschätzungen durch Expertinnen und Experten vorliegen als für die großen Parteien, wird in beiden Befragungen sehr ähnlich eingeschätzt.

Abbildung 3. Zusammenhang zwischen der Links-Rechts Dimensionen im Polit-Kompass (Y-Achse) und CHES (X-Achse).

Im nächsten Schritt wird die Einordnung auf der Populismus-Skala verglichen (Abbildung 4). Ein direkter Vergleich der Polit-Kompass-Werte mit dem Populismusgrad aus POPPA zeigt zunächst, dass Parteien sich in den Quadranten links unten oder rechts oben befinden. Das bedeutet, dass alle Parteien, die bei POPPA einen Wert von über 50/100 erreicht haben, auch im Polit-Kompass mit einen Populismus-Wert von über 50/100 eingestuft sind (und umgekehrt). Mit anderen Worten, wenn die Skalenmitte als Grenze für die Klassifizierung als populistische bzw. nicht-populistische Partei herangezogen wird, kommen beide Befragungen zum gleichen Ergebnis. Dies gilt auch für die binäre Einordnungen in der PopuList, die farblich hervorgehoben sind. In der Grafik rot markierte Parteien werden dort als nicht-populistisch, blaue markierte Parteien (Linke und AfD) als populistisch verortet. Sowohl Die Linke als auch die AfD werden auch im Polit-Kompass oberhalb des Mittelwertes platziert. Die große Gemeinsamkeit ist beachtlich, da die drei Befragungen methodisch sehr unterschiedlich vorgehen (Aggregierung auf der Basis von Survey Items im Polit-Kompass, gewichtete Aggregierung von Subdimensionen bei POPPA, binäre Klassifizierung in der PopuList). Der innerhalb der linkspopulistischen Parteienfamilie eher geringe Populismusgrad der Linken wurde in der vergleichenden Parteienforschung immer wieder diskutiert (Hough & Keith, 2019; March, 2011; March & Mudde, 2005). Auf Seiten der AfD besteht spätestens seit dem Ausscheiden von Parteigründer Lucke innerhalb der vergleichenden Populismusforschung kaum noch Zweifel, dass es sich um eine klassische rechtspopulistische Partei handelt, die in anderen europäischen Ländern bereits seit vielen Jahren erfolgreich agieren (Häusler, 2018; Heinisch & Werner, 2019; Lewandowsky et al., 2016).

Einige Parteien sind im Polit-Kompass etwas populistischer eingeordnet als bei POPPA, so die Grünen, FDP, im geringeren Maße auch CDU und SPD. Dies ist vor allem auf die Auswahl der Populismus-Items im Polit-Kompass zurückzuführen, die auf Survey Items basieren (Castanho Silva et al., 2020). Durch die Ergänzung um das Thema Volksentscheide auf Bundesebene kommt der Volkszentrierung (People-Centrism) ein stärkeres Gewicht zu. Diese Dimension ist eine notwendige, aber keine hinreichende Bedingung für Populismus, da diese auch unter vielen größeren Parteien verbreitet ist (Jagers & Walgrave, 2007; Lewandowsky et al., 2016; March, 2017). Die Variable „Peoplecentrism“ bei POPPA ist hingegen schon so konstruiert, dass sie den konkreten Gegensatz zwischen Bevölkerung und Elite aufgreift.1 Entsprechend fallen also die Einordnungen auf dieser Subdimension für nicht-populistische Parteien bei POPPA niedriger aus. Hier zeigt sich auch der Zielkonflikt bei der Konstruktion von politischen Landschaften für Voting Advice Applications, da manche Items auf der Ebene der Wahlberechtigten möglicherweise anders skalieren als auf der Parteiebene (Mendez, 2014).

Insgesamt zeigt sich also, dass sich trotz der sehr unterschiedlichen Ansätze eine große Übereinstimmung zu anderen politikwissenschaftlichen Befragungen unter Expertinnen und Expert ergibt. Zusätzlich werden im Polit-Kompass aber auch etablierte Kleinstparteien wie die Piraten, die PARTEI, die Freien Wähler oder Volt verortet, die in den meisten anderen Befragung nicht positioniert wurden.

Abbildung 4. Zusammenhang zwischen Populismus-Dimensionen im Polit-Kompass (Y-Achse), POPPA (X-Achse) und PopuList (rot=nicht populistisch, blau=populistisch).




Der Polit-Kompass stellt eine neue VAA dar, die auch unabhängig von Wahlen bei der politischen Orientierung helfen soll und innovative politikwissenschaftliche Methoden kombiniert. Die Parteipositionen wurden dabei mithilfe eines Expert Surveys erhoben. Dabei verortet er die Nutzerinnen und Nutzer in einem zweidimensionalen Raum: auf einer Links-Rechts-Achse, die wirtschafts- und gesellschaftspolitische Konflikte abbildet, und einer Populismus-Achse, die in den letzten Jahren stark an Bedeutung gewonnen hat. Die Entwicklung des Polit-Kompasses, die im Rahmen des Projekts PRECEDE erfolgt, ist dabei eng in die politikwissenschaftliche Forschung eingebettet und basiert auf der Analyse von VAA-, Partei- und Individualdaten. Die Parteien werden durch eine innovative Methode auf der Populismus-Dimension verortet, indem Parteien entlang validierter Populismus Survey Items platziert werden. So können Positionen von Parteien und ihren (potenziellen) Wählerinnen und Wähler direkt miteinander verglichen werden. Die Ergebnisse sind insgesamt valide, was der Vergleich zu anderen Projekten wie der CHES, der PopuList oder der POPPA zeigt – trotz unterschiedlicher methodischer Ansätze in der Verortung.


Bakker, R., Hooghe, L., Jolly, S., Marks, G., Polk, J., Rovny, J., Steenbergen, M., & Vachudova, M. A. (2019). 2019 Chapel Hill Expert Survey. Version 2019.1. University of North Carolina, Chapel Hill. chesdata.eu

Castanho Silva, B., Jungkunz, S., Helbling, M., & Littvay, L. (2020). An Empirical Comparison of Seven Populist Attitudes Scales. Political Research Quarterly, 73(2), 409–424. https://doi.org/10.1177/1065912919833176

Gabel, M. J., & Huber, J. D. (2000). Putting Parties in Their Place: Inferring Party Left-Right Ideological Positions from Party Manifestos Data. American Journal of Political Science, 44(1), 94–103. https://doi.org/10.2307/2669295

Gemenis, K. (2015). An iterative expert survey approach for estimating parties’ policy positions. Quality & Quantity, 49(6), 2291–2306. https://doi.org/10.1007/s11135-014-0109-5

Häusler, A. (Hrsg.). (2018). Völkisch-autoritärer Populismus: Der Rechtsruck in Deutschland und die AfD. VSA-Verlag.

Heinisch, R., & Werner, A. (2019). Who Do Populist Radical Right Parties Stand for? Representative Claims, Claim Acceptance and Descriptive Representation in the Austrian FPÖ and German AfD. Representation, 55(4), 475–492. https://doi.org/10.1080/00344893.2019.1635196

Hough, D., & Keith, D. (2019). The German Left Party. In G. Katsambekis & A. Kioupkiolis, The Populist Radical Left in Europe(S. 129–144). Routledge.

Jagers, J., & Walgrave, S. (2007). Populism as Political Communication Style: An Empirical Study of Political Parties’ Discourse in Belgium. European Journal of Political Research, 46(3), 319–345. https://doi.org/10.1111/j.1475-6765.2006.00690.x

Krouwel, A., Vitiello, T., & Wall, M. (2012). The practicalities of issuing vote advice: A new methodology for profiling and matching. International Journal of Electronic Governance, 5(3/4), 223. https://doi.org/10.1504/IJEG.2012.051308

Ladner, A., & Fivaz, J. (2012). Voting Advice Applications. In N. Kersting, Electronic Democray (S. 177–198). Budrich.

Lewandowsky, M., Giebler, H., & Wagner, A. (2016). Rechtspopulismus in Deutschland. Eine empirische Einordnung der Parteien zur Bundestagswahl 2013 unter besonderer Berücksichtigung der AfD. Politische Vierteljahresschrift, 57(2), 247–275. https://doi.org/10.5771/0032-3470-2016-2-247

Lewis, P., Clarke, S., Barr, C., Kommenda, N., & Holder, J. (2018, November 20). Revealed: One in four Europeans vote populist. The Guardian. http://www.theguardian.com/world/ng-interactive/2018/nov/20/revealed-one-in-four-europeans-vote-populist

Lindstädt, R., Proksch, S.-O., & Slapin, J. B. (2020). When Experts Disagree: Response Aggregation and its Consequences in Expert Surveys. Political Science Research and Methods, 8(3), 580–588. https://doi.org/10.1017/psrm.2018.52

March, L. (2011). Radical left parties in Europe. Routledge.

March, L. (2017). Left and Right Populism Compared: The British Case. The British Journal of Politics and International Relations, 19(2), 282–303. https://doi.org/10.1177/1369148117701753

March, L., & Mudde, C. (2005). What’s Left of the Radical Left? The European Radical Left After 1989: Decline and Mutation. Comparative European Politics, 3(1), 23–49. https://doi.org/10.1057/palgrave.cep.6110052

Marks, G., Hooghe, L., Nelson, M., & Edwards, E. (2006). Party Competition and European Integration in the East and West: Different Structure, Same Causality. Comparative Political Studies, 39(2), 155–175. https://doi.org/10.1177/0010414005281932

Meijers, M. J., & Zaslove, A. (2020). Measuring Populism in Political Parties: Appraisal of a New Approach. Comparative Political Studies, 0010414020938081. https://doi.org/10.1177/0010414020938081

Mendez, F. (2014). What’s Behind a Matching Algorithm? A Critical Assesment of How Voting Advice Applications Produce Voting Recommendations. In D. Garzia & S. Marschall, Matching Voters with Parties and Candidates: Voting Advice Applications in Comparative Perspective (S. 49–66). ECPR Press.

Mudde, C. (2017). Populism: An Ideational Approach. In C. Rovira Kaltwasser, P. A. Taggart, P. Ochoa Espejo, & P. Ostiguy (Hrsg.), The Oxford Handbook of Populism (S. 27–47). Oxford University Press.

Norris, P., & Inglehart, R. (2019). Cultural Backlash: Trump, Brexit, and the Rise of Authoritarian-Populism. Cambridge University Press.

Reiljan, A., Ferreira Da Silva, F., Cicchi, L., Garzia, D., & Trechsel, A. H. (2020). EU Profiler/euandi trend file (2009-2019)[Data set]. https://cadmus.eui.eu//handle/1814/65944

Rooduijn, M., Van Kessel, S., Froio, C., Pirro, A. L. P., de Lange, S., Halikiopoulou, D., Lewis, P., Mudde, C., & Taggart, P. (2020). The PopuList. An Overview of Populist, Far Right, Far Left and Eurosceptic Parties in Europe. Version 2.0. https://popu-list.org/

Schäfer, A., & Zürn, M. (2021). Die demokratische Regression (Originalausgabe Edition). Suhrkamp Verlag.

Thomeczek, J. P. (im Erscheinen). Voting Advice Applications im Kontext der politikwissenschaftlichen Forschung. In N. Kersting, J. Radtke, & S. Baringhorst, Handbuch Digitalisierung und politische Beteiligung. Springer.

Thomeczek, J. P., Jankowski, M., & Krouwel, A. (2019). Die politische Landschaft zur Bundestagswahl 2017. Befunde aus zwei Voting Advice Applications und dem Chapel Hill Expert Survey. In K.-R. Korte & J. Schoofs (Hrsg.), Die Bundestagswahl 2017. Analysen der Wahl-, Parteien-, Kommunikations- und Regierungsforschung (S. 267–291). Springer VS. https://link.springer.com/chapter/10.1007/978-3-658-25050-8_12


Thomeczek, Jan Philipp et al. (2021): Der Polit-Kompass : Eine neue VAA zum aufkeimenden Populismus, Kurzanalyse, Erschienen auf: regierungsforschung.de. Online Verfügbar: https://regierungsforschung.de/der-polit-kompass-eine-neue-vaa-zum-aufkeimenden-populismus/

This work by Jan Philipp Thomeczek, Norbert Kersting, André Krouwel, Susan Banducci, Lorenza Antonucci, Carlo D’Ippoliti, Andrei Zhirnov, Laszlo Horvath and Christian Mongeau is licensed under a CC BY-NC-SA license.

  1. Formulierung im Codebook: „Some parties believe that sovereignty should lie exclusively with the ordinary people (i.e. the ordinary people, not the elites, should have the final say in politics). Please tick the box that best describes the extent to which each party considers the ordinary people to be sovereign.” []

[Blog] Stealth populism: Explaining the rise of the Alliance for the Unity of Romanians

This article was first published on the LSE EUROPPBLOG on Feb 26 2021 (https://blogs.lse.ac.uk/europpblog/2021/02/26/stealth-populism-explaining-the-rise-of-the-alliance-for-the-unity-of-romanians/)

One of the major stories from the 2020 Romanian election was the success of a new nationalist party, the Alliance for the Unity of Romanians, which surprised observers by winning over 9 per cent of the vote. Drawing on new data, Mihnea Stoica, Vladimir Cristea and André Krouwel explain how the party managed to build support so rapidly.

The most recent legislative elections held in Romania in December 2020 produced what was regarded by most political observers as a major surprise: a brand new, almost obscure, populist party, the Alliance for the Unity of Romanians – with an acronym that means ‘gold’ (AUR) – managed to gain over 9 per cent of the vote. The success of this ‘stealth party’ was missed by most pollsters and analysts. In the local elections that were held only two months before the legislative elections, the AUR only gained 1 per cent of the vote.

On top of that, several smaller parties fell victim to the 5% threshold for entering Parliament. The most remarkable losers were the Popular Movement Party (PMP) of former President Traian Băsescu and the PRO Romania Party of former Prime Minister Victor Ponta. The outcome also indicated a fragmented electorate, with no party managing to win more than 30 per cent of the vote. Finally, the incumbent government parties won enough seats to remain in office, which was also a departure from the typical ‘throwing-the-rascals-out’ mentality of voters in Romania.

The Romanian political landscape in the 2020 elections

As part of a recent project, a team of Romanian academics from Babeș-Bolyai University developed a so-called Vote Advice Application (VAA) for the 2020 Romanian elections. This online application allowed users to compare their opinions with those of the main parties. Users answered a set of 30 questions on political issues. Based on these answers, the VAA matched the user’s political preferences with party positions on the same issues.

The parties were coded on the 30 topics according to their official positions, campaign documents and speeches by party leaders. The tool framed the positions of the parties on salient issues so as to give an ideological direction and determine their loading on one of the deeper cleavage dimensions that demarcate the Romanian political landscape, producing a multidimensional ‘map’ of party competition. Users could find where they were positioned in the political space, and how close or far they were from the main political parties. Figure 1 below shows this map of the Romanian political landscape.

Figure 1: Map of the Romanian political landscape

Note: The figure shows parties on a dimension between progressive (liberal) and conservative (conservator) values, and on a dimension between left (stânga) and right (dreapta) positions on the economy. The parties shown in the figure are the Social Democratic Party (PSD – left side of the figure), National Liberal Party (PNL – arrow logo, top right), 2020 USR-PLUS Alliance (blue/red split logo, top left), Alliance for the Unity of Romanians (AUR – bottom left), Democratic Alliance of Hungarians in Romania (UDMR – red logo, top left), Popular Movement Party (PMP – apple logo, right side), and PRO Romania (blue/red/yellow logo, top left).

For the 2020 elections, parties seemed to be mostly crowded on the left side of the political map. The Social Democratic Party (PSD) was the party positioned most to the left, but it was judged to be quite moderate on the progressive-conservative dimension. PRO Romania, which is a splinter party that broke off from the PSD, was also on the left, but with a slightly more progressive platform. The Democratic Alliance of Hungarians in Romania (UDMR), which mainly represents Romanian citizens with a Hungarian ethnic background, occupied a very similar position to PRO Romania.

The USR-PLUS alliance was the most progressive left-wing competitor. This alliance consists of the Save Romania Union (USR) and the Freedom, Unity and Solidarity Party (PLUS), two parties with fairly flexible ideological orientations. The Alliance for the Unity of Romanians ended up being the most conservative left-wing party. Voters that were located on the right side of the map were likely to end up closest to either the National Liberal Party (PNL) – a progressive and pro-European party which was in government at the time of the elections – or the PMP, which is a conservative party that shares some common ground with the Republican Party in the United States.

Stealth populism

By far the most unexpected outcome of the elections was the emergence of the AUR, which established itself as the fourth largest political force in the Romanian Parliament. In recent years, Romania has been one of the few countries in the European Union without a major nationalist-populist political force, following the decline of the Greater Romania Party (PRM) at the beginning of the 2010s.

What made the electoral result of the AUR particularly surprising for political observers and the general public alike is that the party did not even appear in the opinion polls conducted before the election. Polls only started to include it in the last days before the election, and even in these cases its score was massively underestimated. Yet while many Romanians heard about the AUR for the first time on the night of the elections as the exit-polls were announced, the party had been gradually growing their visibility throughout 2020, mostly through an efficient and well-targeted campaign on social media.

Most of the party’s main positions focus on Romanian nationalism, conservatism and protectionism. According to the party’s political programme, the four main pillars of the AUR’s ideology are “family, fatherland, faith and freedom”, and their campaign efforts commonly featured activism for the unification between Romania and Moldova, an opposition to progressive politics, criticism of foreign investment, and scepticism towards further European integration.

To sketch a profile of the AUR’s supporters, we use data collected during the election campaign – through the VAA as well as a post-election survey on a panel that is weighted to be representative for the total population. Respondents were asked who they voted for, but also to estimate the likelihood that they would ever vote for each of the other main parties competing in the election. This ‘propensity to vote’ (PTV) measure ranged from 0 (highly unlikely) to 10 (highly likely). We also asked voters to place themselves on a standard left-right scale, from 0 to 10.

If we regress vote propensities of the AUR against the self-declared left-right dimension we get no significant result. A regression of vote propensities for the party against the calculated left-right dimension based on all substantive statements on the economic dimension indicates that AUR voters tend to lean to the right in policy preferences. However, when we only take voters with a high propensity to vote for the party (greater than or equal to 6 on the scale), this effect is no longer significant.

Voters for the AUR are not very pronounced on the left-right dimension, but are very mixed both in self-declared ideological orientation as well as actual policy preferences. The appeal of the AUR is on the cultural dimension, where AUR voters are much more pronounced. In their conservative-progressive orientation, AUR voters are clearly conservative, for example opposing abortion rights, same-sex marriage and immigration. This finding matches with the position of the AUR in the Romanian political landscape, based on its political programme, where the party displays a mix of deeply conservative stances on the cultural dimension, but is moderately left-wing on the economic dimension (see Figure 1).

On economic issues, AUR voters seem to be highly diverse in terms of the left-right spectrum, and voter opinions on substantive issues show conflicting responses. For example, vote propensities for the AUR are essentially unrelated to responses to the statement that “the government should offer unemployed people more financial support”, yet they favour wealth redistribution from the rich to the poor. However, AUR supporters are also more likely to be opposed to government intervention in the economy and agree with decreases in government spending. When we plot such contradictory economic mindsets, we get the results shown in Figure 2 below.

Figure 2: Political positions of Romanian voters (AUR voters highlighted)

Note: The figure shows an estimate of the political positions of all voters contained in the sample, based on their responses to questions asked via the Vote Advice Application and a separate survey.

On European integration, AUR voters appear to be much more critical of the European Union than the general electorate. They are more likely to agree with the statement that “overall, EU member status is a bad thing for Romania” and that “European decision-makers are treating Romania like a colony”. They also tend to agree with the prioritisation of Romanian investors over foreign ones when awarding governmental aid and oppose the right of Hungarian-Romanians to education in their native language.

Covid-19, climate change, and corruption

Besides these conventional topics, a major reason for the success of the AUR’s campaign was the position many of the party’s prominent figures took on the Covid-19 pandemic. Throughout the outbreak, they criticised the government’s response, called for a loosening of lockdown measures, campaigned against face mask rules, played down the severity of the virus, organised gatherings and rallies without observing social distancing rules, and in some cases even promoted conspiracy theories about the virus or vaccines. In a country where 22% of the population does not want to take a Covid-19 vaccine and another 9% claims to be unlikely to take it, these actions gave the party widespread visibility and attracted sympathy from some parts of society.

These attitudes can also be seen in the data, as a preference for the AUR is positively associated with a belief that “the government should allow people to make their own decisions on how they want to protect themselves from the coronavirus during this pandemic”, and that “churches should remain open during the pandemic”. With respect to another topic where populist parties have a tendency to disagree with the expert consensus, climate change, the AUR fits the prevailing trend as their voters are more prone to believe that “the so-called climate crisis is an exaggeration” than the voters of other parties.

Another core message of the AUR campaign was its fight against government corruption and the political establishment. Analysing the data, it becomes clear how these messages have resonated so strongly, as AUR sympathisers appear to be less trusting of Romanian institutions than the voters of any other party. “The government is led by interest groups that aim to protect themselves” and “quite a lot of the people leading the government are corrupt” are two statements that they strongly resonated with, and AUR voters also seem to be the most pessimistic about the direction in which Romania is currently heading.

Sources of support

After the unexpected result in December, political analysts began to assess where AUR voters had switched their allegiances from. Our data reveals that there is no one party from which AUR has won its voters. Instead, their voters include people who voted for all of the other parties in 2016, as well as a very substantive proportion of people (as high as 50% of their electorate) who did not vote in 2016, were not allowed to vote because they were under the legal voting age, or voted for smaller fringe parties that did not enter the Parliament.

Demographically, our data shows that AUR voters are more likely to be male and have a lower level of education. Interestingly, compared to parties such as USR-PLUS or the PSD, whose electorates are very age-specific (youth in the case of the former, the elderly in the case of the latter), the AUR seems to attract voters of all ages. Nevertheless, the party is marginally more popular among younger people. These findings are consistent with previous results presented by CURS-Avangarde’s exit-poll.

While nationalist political mobilisation is no stranger to Romanian politics, as illustrated by the previous success of the Greater Romania Party, the rise of the AUR from an almost insignificant force to the fourth largest party in Parliament was undoubtedly surprising. Many Central and Eastern European countries have experienced an illiberal, populist and ultraconservative wave over recent years, so there are reasons to believe the AUR may be here to stay. However, whether the party will be able to consolidate their rather erratic and inconsistent voter base remains to be seen.