The emergence of radical right parties in European countries and the influence of electoral systems on this process

(1)

The emergence of radical right parties in European countries

and the influence of electoral systems on this process

Master Thesis

Student: Carien van der Poel Student number: 1254979 Supervisor: Dr. A. Afonso

Master’s Programme in Public Administration (Specialisation: Economics & Governance) Leiden University

(2)

2 Abstract

This thesis studies the influence of electoral systems on the electoral strength of radical right parties in Western European countries. To test the established hypotheses, a nested analysis is applied. The nested analysis consists of a statistical analysis of a large number of cases and a more intensive investigation of four cases (France, UK, Austria and the Netherlands). The electoral systems have a significant influence on the electoral strength of radical right parties. Radical right parties have the best opportunities to flourish in a proportional representation system compared to a majoritarian system. An increase in the proportionality of the electoral systems results in an increase of the electoral strength of radical right parties. Based on the statistical analysis, Duverger’s law concerning the strategic voting motives is refuted.

(3)

3

Illustrations

Figures

2.1: Effective number of legislative parties P. 10

3.1: The relation between the total seats of radical right parties P. 23 and the index of absolute disproportionality, per country

4.1: Trend total votes of radical right parties between 1980-2015 P. 24 4.2: Geographical concentration of PVV supporters 2015 P. 39 4.3: National election results of the United Kingdom by district, 2017 P. 53 8.1: Party positioning of French political parties on different themes P. 70 8.2: Party positioning of Dutch political parties on different themes P. 71 8.3: Party positioning of Austrian political parties on different themes P. 72 8.4: Party positioning of United Kingdom’s political parties on different themes P. 73

Tables

2.1: Effect of electoral systems on the success of radical right parties P. 10

2.2: Turkish election results 2002 P. 13

4.1: Total right seats and total rights votes by electoral system P. 25 4.2: The difference in total right votes by electoral system P. 25 4.3: Correlation between the total votes and total seats for radical right parties P. 26 4.4: Regressions on the relation between total votes and total seats of radical P. 26 right parties

4.5: Regressions on the relation between the total seats of radical right parties and P. 27 index of absolute disproportionality

4.6: Regressions on the relation between total right seats and the effective number P. 28 of parties

4.7: Overview of the President in France since 1958 P. 30

4.8: Election results of the French Presidential elections in 2017 P. 32 4.9: Election results of the Dutch Parliamentary elections in 2017 P. 37

4.10: Austrian Cabinets since 1983 P. 43

4.11: Election results of the Austrian national parliamentary elections in 2017 P. 45

4.12: British Cabinets since 1970 P. 49

4.13: Results of the elections of the House of Commons in 2017 P. 52

5.1: Results of the national elections in 2017 P. 56

8.1: Variable descriptions P. 67

8.2: Summary statistics P. 67

8.3: Average number of votes and seats for right parties from 1980 till 2015 P. 68 8.4: Variable descriptions of Chapel Hill Expert Survey (1999-2014) P. 69

(5)

5

Chapter 1: Introduction

After years of dominance of the established parties, political stability and maybe even monotony, the political spectrum is subject to change. The established political parties can no longer automatically count on their electorate and the number of undecided voters increases every year. New norms are valued, such as freedom of speech and the environment, instead of traditional thoughts as religion and family values. This new thinking gives room for new parties to make their entrance in the political field in European countries.

The increase in interest in new standards and values is also noticeable in the rise of radical right parties in Europe. The anti-immigrant feelings, the anxiety of losing their own culture and living standards and anti-European Union beliefs are merged into several radical right parties in Europe. Although the focus of these parties differs, their footing corresponds. Examples of radical right parties are Front National in France, PVV in the Netherlands and Fronte Sociale Nazionale in Italy. The radical right is the first family, besides the traditional parties, who received so many votes since the Second World War. The political field has changed fundamentally and this also influenced the strategy of the established parties who lost great amounts of votes.

Although many Europeans have witnessed the rise of a radical right party, the representation of these parties in Parliament differs. This thesis will analyse the emergence of radical right parties. How did the radical right parties receive so many votes during the last years in European countries? Is there an influence of the electoral systems on the existence of radical right parties? According to Duverger, the rise of new parties differs between the two electoral systems. He argues that countries with a proportional representation have multiple parties in Parliament. On the other hand, countries with a majoritarian system will have only two parties in Parliament. Because of their majority rule, third parties will hardly be represented in Parliament. Duverger argues that voting for third parties will be absent, because of the limited chance of representation in Parliament. He argues that voters favour voting for a party that will be present in Parliament, although this party is not closed to their political perspectives. Duverger’s law about the number of parties in Parliament will also influence the rise of radical right parties. Based on this law, it is expected that countries with a proportional representation will have more votes and seats for radical right parties, because of their increased chance of governing. On other hand, countries with a majoritarian system will have no or a limited number of votes and seats for radical right parties because of the meager opportunity to be represented in Government.

To examine the emergence of radical right parties and the influence of electoral systems on this process, the research consists of two different approaches. The first part will research the differences between both elections and their influence on the electoral strength of radical right parties based on a statistical analysis of multiple Western European countries. The second part will research this relation based on case studies of France, the Netherlands, Austria and the United Kingdom. The case studies will be focused on the political development of a country

(6)

6 in which the radical right parties emerged. The characteristics of a country’s election system are defined and the election results of the last decades are analysed.

Based on the two approaches the following research question will be answered: What is the

influence of electoral systems on the electoral strength of radical right parties?

Sub-questions which will be answered:

- What is the relative difference in votes of radical right parties in the proportional representation and the majoritarian systems?

- Do radical right parties have relatively more votes and seats in proportional systems then in majoritarian systems?

- How did the mainstream parties respond to the emergence of radical right parties?

1.1: Contribution to the literature Previous research has shown that the electoral system is one of the determining factors of the number of parties in Parliament. There is a positive relationship between the proportionality of the electoral system and the numbers of political parties. The more proportional an election system, the more parties presented in Parliament. A comparison between the United Kingdom and the Netherlands confirms this relationship. The United Kingdom with a majoritarian system is characterized by two major parties, namely the Conservatives and the Labour Party. While the Netherlands with a proportional representation is characterized with multiple parties in Parliament and Government.

Although the research about radical right has increased, it is still not clear why people vote for radical right parties and how these parties emerged. Based on the existing literature, it could be expected that the number of votes for radical right parties would be very modest in a majoritarian system because of the limited chance of governing. The radical right parties are the weak third party, which will have a hard time collecting a majority of votes due to their extreme views. Therefore, we can expect that radical right parties have no reason to exist in the majoritarian system because people will give their votes to a stronger party with a chance of winning. As we have seen in France, the radical right parties have collected a great amount of votes, although it was a majoritarian system. This thesis will test if Duverger’s law also counts for radical right parties. Does the proportionality of the system effects the electoral strength of radical right parties? And is there a difference between the proportional representation and the majoritarian system?

According to Norris, the electoral strength between both systems differs due to the allocation of seats. Although the number of votes in both systems is comparable, the number of seats differs significantly. He argues that radical right parties have the best chance of governing in a proportional representation. This conclusion is based on data of the elections between 1990 and 2004. As mentioned before, the existence of radical right parties in European countries has increased enormously. The prominence of the radical right parties over the last years asks for new research to determine the influence of electoral systems on the strength of radical right parties. This thesis will analyse the most recent elections of Austria, France, the Netherlands and the UK. All of these elections were held in 2017. The influence of the electoral systems on the electoral strength of radical right parties of these elections has not

(7)

7 been researched before. The actuality of the results contributes to the established research as the electoral strength of radical right parties in Western Europe has strongly increased. The research consists of two approaches, a statistical analysis of 15 countries and an in-depth case study of four countries. Combining these approaches improves the robustness of the results and conclusions. Based on the results of the Large N analysis and the context-based analysis of the case studies, it can be concluded that electoral systems have a significantly different influence on the electoral strength of radical right parties. The influence of the electoral system will be further discussed in the next chapters.

1.2: Reading guide

This thesis proceeds as follows. The following chapter presents the established literature concerning electoral systems, electoral thresholds and the definitions of the radical right parties. The chapter ends with the testable hypotheses. Chapter three describes the research design, defining the methods and data. Chapter four presents the results. These are divided in the results of the statistical analysis and the results of the case studies. Chapter five offers the discussion and the last section concludes and answers the research question.

(8)

8

Chapter 2: Literature review

The following chapter will discuss the established literature on the definitions of electoral systems, thresholds and their influence on new minor parties and radical right parties. This overview is followed by an elaboration of the characteristics of radical right parties and the explanations of their emergence in the last decades.

2.1: Electoral systems

In a democracy, voters give their voices through an election of representatives. The election of these representatives is an important part of a democracy because it builds the connection between the wishes of citizens and the behaviour of policymakers (Powel, 2000: 14). The way the Parliament is elected is called the electoral system. Following Lijphart, there are two broad electoral systems, the majoritarian system and the proportional representation (Lijphart, 2012: 10). In a majoritarian system or a Westminster model, seats are awarded to those candidates who obtained a plurality of votes (Lijphart, 2012: 10). On the other hand, the proportional representation or consensus democracy is an electoral system in which nearly all votes count towards the final proportions in the distribution of seats (Lijphart, 2012: 30). The main difference between both systems is the representation of public interests in government. In a majoritarian system, the government is a representation of the majority via single-party executives, while the consensus democracy is focused on including as much of interests as possible through the distribution of power (Bormann, 2010: 2).

The electoral systems influence the number of political parties in parliament and government in a country (Lijphart, 2012: 61). According to Duverger’s law “the simple majority single-ballot system favours the two-party system and the simple-majority system with second single-ballot and proportional representation favours multi-partyism" (Duverger, 1972). In a majoritarian system, there will be two parties in parliament, while in a consensus democracy there are multiple parties represented in parliament and government. Third parties are pushed away in a majoritarian system due to the distribution of seats. Only the winner of a district will gain a seat and the party that comes third will gain nothing even if they received a substantial amount of the votes (Duverger, 1972). Converting the votes into seats is not based on an equal distribution and especially third parties are underrepresented. In this system, a third party can receive ten percent of the total votes, but only have one percent of the seats. This is certainly the case if the votes for the third party are geographically dispersed (Duverger, 1972).

Duverger states that there are two reasons for the absence of third parties in majoritarian systems. First, third parties tend to aggregate with others to increase the chance of winning (Duverger, 1972). Second, the influence of tactical voting. Voters are more likely to desert the weaker party and support a strong party which has the opportunity to govern (Duverger, 1972). Voters make a consideration between voting for a third party which is closest to their personal interests but with the chance of wasting their vote, and choosing between one of the two strong parties which is not the closest to their personal interests but which has the certainty of governing (Norris, 2004: 109). If voters have a specific preference for a party and the ability to estimate the chance of winning, they will vote strategic (Norris, 2004: 109). Minor parties lose votes due to this strategic voting based on a cost-benefit analysis (Norris,

(9)

9 2004: 108). The rise of new parties is influenced by multiple opponents, such as funding access, but a country’s electoral system is the main determinant of the rise of new parties (Norris, 2004: 105). Majoritarian systems try to limit the opportunities for new parties (Norris, 2004: 109). But the upswing is not insurmountable (Duverger, 1972). A third party can enter the political field if it uses the shortcoming of the two strongest parties (Duverger, 1972).

In a proportional representation, third parties have a great opportunity to be elected and represented in parliament because of the equal distribution of seats (Duverger, 1972). In a pure proportional representation, parties do not have to unite and voters do not have to vote strategic for the strongest parties to be represented in parliament, as all parties have the same chance of achieving a seat (Duverger, 1972). Proportional representation systems are therefore more competitive and fragmented (Norris, 2004: 109).

Arend Lijphart supports Duverger's law stating: “plurality electoral systems tend towards party dualism and proportional representation to multipartyism” (Duverger, 1972). A more recent study by Norris based on the elections of 1995 until 2000, found similar results. Countries with a proportional representation have twice as many parties in parliament compared to countries with a majoritarian system (Norris, 2004: 112).

Another way of estimating the fragmentation of the election is by the index of the effective number of parties established by Laakso and Taagepara (Laakso & Taagepara, 1979: 4). The effective number of parties is a measure of the number and relative strength of the political parties in a country’s electoral system (Laakso & Taagepara, 1979: 4). Laakso and Taagepara make a distinction between the votes shares of parties, which is called the effective number of elective parties, and the seat shares in parliament, which is called the effective number of legislative parties (Laakso & Taagepara, 1979: 4). The effective number of parties gives an index, which indicates the number of parties in a country if all parties are equally sized (Laakso & Taagepara, 1979: 5). An index of five is equal to a fragmentation of five parties of equal size. This index makes it possible to compare the party competition in different countries.

(10)

10 Figure 2.1: Effective number of legislative parties

Source: The database of WHO governs in Europe and beyond. At: https://whogoverns.eu/party-systems/effective-number-of-parties/

This figure displays the index of the effective number of parties based on the number of seats. Countries with a proportional system have a higher index, which indicates a higher level of party fragmentation. For example, Germany and the Netherlands with an index of approximately five. The fragmentation in countries with a majoritarian system is lower. The UK and France have both an index around two, which indicates a party system with two equal-sized parties.

Majoritarian and proportional representation systems have a different effect on the rise of new parties. This also counts for the number of radical right parties in a country. Norris researched the effect of electoral systems on the electoral fortunes of radical right parties (Norris, 2004: 109).

Table 2.1: Effect of electoral systems on the success of radical right parties

Source: Norris, P. (2004) Radical Right: Voters and Parties in the Electoral Market. (CUP, Cambridge 2006). P. 112.

The table above shows the results of the electoral fortunes of radical right parties. The number of votes for radical right parties was in both systems almost equal with the most recent elections, 7.1 percent in majoritarian systems and 7,2 percent in proportional electoral systems

(11)

11 (Norris, 2004: 112). There are two explanations for this similarity. First, voters do not vote tactical. The costs of voting for another party, which has the possibility to govern, are too high. These parties are “too far away from their ideal policy position” (Norris, 2004: 113). Therefore, the voters chose the radical right party even if the chance of governing is minimal. Second, voters might use their vote more symbolic instead of rational (Norris, 2004: 113). By supporting the radical right, voters tried to influence the future policy indirectly, even if the right parties did not gain any seats. Due to this symbolic vote, governing parties might adapt their policy (Norris, 2004: 113).

However, the similar percentage of votes did not result in a similar percentage of seats in both systems. The radical right parties gained 3.8 percent of the seats in majoritarian systems and 9.9 percent in proportional representations (Norris, 2004: 114). Based on this evidence, there is a significant effect of electoral systems on the rise of extreme right parties where extreme right parties have the best possibility to gain seats in proportional representation. The limitations of radical right parties to flourish in majoritarian systems is not because of tactical voting but due to the distribution of seats (Norris, 2004: 114).

To conclude, new radical right parties have the best opportunities to evolve if there are low barriers (Ignazi, 2003: 7). This is the case in a proportional representation system. Due to the different allocation of seats, majoritarian systems have automatically a threshold to hinder party competition (Norris, 2004: 107). Proportional representation systems, on the other hand, are associated with multiple small parties and more extreme ideologies (Norris, 2004: 107). Countries with a proportional representation might introduce electoral rules to hinder party competition. The following section will further elaborate on these electoral tresholds.

2.2: Electoral threshold

Countries have the possibility to introduce multiple thresholds to exclude minor parties from parliament (European Commission, 2008: 2). As mentioned in the previous part, the electoral system itself is a threshold which influences party fragmentation. A majoritarian system is better capable to exclude minor parties, but proportional representations have the possibility to introduce electoral rules to achieve a similar effect. A tool that proportional representation applies is the electoral threshold. An electoral threshold can be defined as “the minimum level of support which a party needs in order to gain representation” (Lijphart, 1994: 11). This can be an absolute number of votes or a percentage of the total number of votes cast (Lijphart, 1994: 11).

There are two types of thresholds to exclude parties, an informal and formal threshold (Lijphart, 1994: 25). Every country has an informal threshold. An informal threshold is defined as “the percentage of votes needed to get one seat at a district level” (European Commission, 2008: 8). The level of this threshold is based on the total number of legislators per constituency, which is called the mean district magnitude (Norris, 2004: 119). This number differs strongly between countries, for example, England with one legislator per district and 150 legislators in the Netherlands (Norris, 2004: 119).

Lijphart has established the following formula to measure the informal threshold: threshold = 75% / (district magnitude + 1) (Lijphart, 2002: 140). The formula emphasises the effect of the

(12)

12 district magnitude on the accessibility of the electoral system (Lijphart, 2002: 140). A low district magnitude has the same impact as a higher formal threshold. If a district magnitude consists of nine seats, the informal threshold is 75/10 = 7.5%. Countries with a limited number of legislators per district have a high threshold for small parties, because a great number of votes is required to win a seat (European Commission, 2008: 8). If the number of legislators is greater, small parties have an increased chance of being elected because of the lower number of votes required. The informal threshold has the same impact on the presence of small parties in parliament as the legal threshold. Other elements that influence the level of the informal threshold are the electoral formula and the number of competing parties (Norris, 2004: 119).

The formal threshold is the minimum percentage of the total votes that a political party must have obtained in order to get a seat (Lijphart, 1994: 25). Examples are Germany with a threshold of five percent and Sweden with a four percent formal threshold (Norris, 2014: 119). The threshold can be implemented at different levels, for example at the national or regional level.

The informal and formal threshold both influence the access for new parties, but the remaining part of this thesis will be focused on the formal electoral threshold.

There is an ongoing debate about the advantages and disadvantages of a formal electoral threshold and countries adjusted their thresholds because of dubiety (Baskaran & Fonseca, 2015: 44). The debate is organized along two sides: legislative cohesion and political representation (Norris, 2005: 120). Countries decide to introduce a proportional representation with the aim to increase proportionality and to accomplish a representation of minorities in politics (Lijphart, 1994: 10). A proportional representation, without an electoral threshold, gives even the smallest minorities the possibility to translate their interests into policies by establishing a political party (Lijphart, 2012: 30). The benefit of proportional representation is that the entire population is represented in the political system.

Supporters of an electoral threshold, on the other hand, emphasize the disadvantages of this party competition for the governability of a country. The increased number of parties in parliament may lead to fragmentation, unstable governments and even state failure (Norris, 2004: 106). Introducing an electoral threshold is therefore aimed at reducing party fragmentation by excluding outliers (Lijphart, 2012: 140). The introduction of a threshold may also encourage faster formation processes (Norris, 2004: 118).

A study by Baskaran and Fonseca shows that abolishing a threshold influences the party competition in a country (Baskaran & Fonseca, 2015: 44). Their study is based on a difference-in-difference strategy of 426 municipalities in Germany (Baskaran & Fonseca, 2015: 49). Municipalities who abolished the threshold are compared with municipalities who maintained the threshold (Baskaran & Fonseca, 2015: 44). By focusing on the cut-off just before the abolishment of the legal threshold and just after the abolishment, the effect of the reform is estimated (Baskaran & Fonseca, 2015: 44). The authors found that the abolishment did not had an effect on voter’s turnout but did influence the distribution of seats (Baskaran & Fonseca, 2015: 60). Abolishing the threshold led to an increase of seat shares for minor

(13)

13 parties, which can be explained due to different voting patterns (Baskaran & Fonseca, 2015: 60). Voters decided to give their votes to minor parties instead of the larger parties. The number of votes for medium parties did not significantly change (Baskaran & Fonseca, 2015: 60).

An electoral threshold results in under-representation and over-representation. Due to the threshold, minor parties disappear from parliament (Lijphart, 2002: 140). A part of the electorate loses its representatives and their interest will not be heard (Norris, 2004: 108). The group is underrepresented in parliament and the distance between them and the politicians will increase if this happens systematically. Excluding minor parties might, therefore, lead to the decline of trust in the democracy (Norris, 2004: 108). Due to the threshold, the larger parties are overrepresented in parliament (Lijphart, 2002: 140). The seats in parliament are divided between fewer parties which will lead to an un-proportional distribution of seats (Lijphart, 2002: 140).

The problem of under- and overrepresentation is illustrated in the following table. Table 2.2: Turkish election results 2002

Source: European Commission (2010). Report on thresholds and other features of electoral system which bar parties from access to parliament (II). P. 6.

This table displays the election results of Turkey in 2002 with a ten percent formal electoral threshold introduced by the established party. The aim of introducing this threshold was to exclude the Kurdish party, a religious minority in Turkey (European Commission, 2010: 5). Due to the electoral threshold, 46.3 percent of the votes are not converted into seats. Almost half of the population is not represented in parliament. The percentage of seats of the established parties doubles compared to the percentage of votes they have received (European Commission, 2010: 6). This over-representation of the main parties leads to an unbalanced representation of interests in parliament. The large established parties benefit from an electoral threshold, because of the larger number of seats and the increased chance of existence and governing.

Small parties are more disadvantaged by an electoral threshold than large parties. This also counts for radical right parties. Norris was interested in the differences of the success of radical right parties in countries with a proportional representation electoral system. Based on the election results of 1990 until 2004, Norris found a significant negative effect between formal electoral thresholds and the seats for radical right parties (Norris, 2004: 121). Countries without an electoral threshold have more radical right parties in Parliament than countries with an electoral threshold (Norris, 2004: 121). The number of votes for radical right parties does not differ between countries with and without an electoral threshold (Norris,

(14)

14 2004: 121). The success of radical right parties in proportional representation without an electoral threshold is because of the possibility to gain enough votes to get a seat in Parliament. From that moment onwards, these parties had the possibility to expand further (Norris, 2004: 121).

The above-mentioned literature explains the influence of electoral systems on the number of parties in a country. Besides the influence on the number of parties, the electoral system also determines the composition of a country’s coalition (Iversen & Soskice: 2006: 165). Iversen and Soskice argue that the coalitions between both electoral systems differ significantly. Countries with a majoritarian are characterised by a center-right coalition, while countries with a proportional representation have on average more center-left coalitions (Iversen & Soskice: 2006: 177). The authors emphasize the influence of classes in determining the coalition with a dominant role of the middle-class party (Iversen & Soskice: 2006: 166). Based on the established literature it is possible to conclude that electoral systems have a significant impact on a country’s political field. It influences not only the number of parties but also the governing coalition. The following research determines the influence of the electoral system on the electoral strength of radical right parties.

2.3: The radical right parties

During the last decades, the political system has changed. New norms determined the political preference, the rise of the Green parties and the emergence of the radical right changed the political party composition (Ignazi, 1995:2). The following part will be focused on the definition of radical right parties and the explanations for their emergence.

Several authors tried to define the family to which radical right parties belong (Mudde 2011, Fennema 1997 & Rydgren 2007). Not only determining the party identification is subject to discussion, authors do not agree on the name for these parties. According to Ignazi, the parties should be called extreme right parties (Ignazi, 2003: 28). He states that the term radical right is too broad to refer only to political parties, it could also refer to other movements than parties with different identities and views (Ignazi, 2003: 28). Another term used for radical right parties is populism (Betz, 1994: 4). Betz introduced this term to not only emphasize the radical views of these parties but also their ability to respond to the fear and dissatisfaction of the common man (Betz, 1004: 4). Other terms to define these parties are anti-immigrant parties, new right parties and far-right parties (Ignazi 1995, Van der Brug & Fennema 2006). To capture a broad range of views in one term results in discord and is not accepted by all scholars. Although there are some downsides, the term radical right parties will be used, as this is the most common term used in the academic discourse (Van der Brug & Fennema, 2006:1).

Radical right parties are also defined in multiple definitions. Even though there is no single definition, the authors focus on the same characteristics of radical right parties. Overlapping features are fascism, populism, nationalism, racism and critical view towards democracy. The fascist ideology has been an inspiration for radical right politicians and Fennema argues that it constitutes the ideological foundation of radical right parties (Fennema, 2006: 7). Fascistic traits such as nationalism, racism, anti-materialism and ethnocentrism are in some way taken

(15)

15 over by radical right parties (Fennema, 2006: 7). Rydgren also affirms the fascist views of radical right parties (Rydgren: 2007: 246). Especially the pressure for change features both ideologies in which the fascist desire a new future and radical rights longing the past (Rydgren: 2007: 246). Radical right parties are therefore frequently referred to as neo-fascist (Ignazi, 1995: 5).

Core of the radical right identity is populism (Golder. 2016: 479). According to their ideology, there is a division between “the pure people and the corrupt elite” (Mudde, 2012: 12). The pure people are the normal citizens of society and the corrupt elite refers to the established political parties but also the media (Golder, 2016: 479). Radical right parties strongly believe in the power of the pure people and feel the need to protect them against the corrupt elites (Golder, 2016: 479). The elites abuse their power by prioritizing international activities and self-interests instead of national and general interests (Rydgren, 2007: 242). The corrupt elite is being accused of keeping power away from the citizens and therefore radical right parties propose multiple measures to increase the influence of the pure people in politics (Golder, 2016: 479).

The radical right parties have a critical attitude towards the democracy (Fennema, 2006: 9). As mentioned in the previous paragraph, radical right parties accuse politicians of selfishness. They do not represent the people or political ideas, they use democracy to implement their own interests (Fennema, 2006: 9). According to Ignazi, all radical right parties share this anti-system attitude (Ignazi, 1995: 5). Although they undermine democracy, the radical right parties do not want to abolish democracy. According to these parties, democracy should be more democratic by introducing a more direct democracy (Fennema, 2006: 10).

Associated with populism, nationalism also a typical determinant for their acts. According to radical right parties, the ethnic nation is paramount and individuals are inferior to the nation (Rydgren, 2007: 242). There is a clear distinction between the members of the nation and those who do not belong to the nation. Radical right parties emphasize the importance of a homogenous nation which generates an exclusion of those who do not belong to the nation (Fennema, 2006: 8). The glorification of the nation legitimizes their xenophobia (Golder, 2016: 480). Others than the nation, threaten the democracy. Radical right parties feel obliged to protect their own people by returning to traditional values and prioritizing national interests (Rydgren, 2007: 242). International interest should also be subordinate to the national significance. The racist stance explains the term anti-immigrant parties (Golder, 2016: 480). The mentioned components of radical right parties can be summarized in a concrete definition of radical right parties, provided by professor Mudde. According to Mudde radical right parties can be defined based on three characteristics: nativism, authoritarianism and populism (Mudde, 2011: 12). Nativism is the belief that the state should only consist of the indigenous population and should protect itself from foreign influences especially of non-native people (Mudde. 2011: 12). Besides that, the society should be well ordered and politicians should solely represent the general will (Mudde. 2011: 12). Only parties who are characterised by all three features can be called radical right parties and are part of the political family (Mudde, 2011: 12).

(16)

16

2.4: Explanations of the emergence of radical right parties The emergence of radical right parties depends on the country, but there are several general

explanations which explain the rise of the radical right parties. The following explanations are not all necessary and sufficient but have contributed to the growing adherents for these parties. The causes can be divided into demand and supply explanations (Rydgren, 2007: 247). The demand explanations are focused on the voters and the supply explanations on the party and party system (Rydgren, 2007: 247). First, the demand explanations will be discussed followed by the supply explanations.

The political system has been very stable for many years. Voting was based on social identity. Every voter voted for the party who belonged to their social group and each party was certain about the number of votes they received. During the last decades, this commitment between voters and parties changed. Due to individualism, the ties between parties and voters are released and party identification is decreased (Ignazi, 1995:2). The number of party members is reduced and voters choose a party based on an individual choice instead of a group process (Ignazi, 1995: 2). Due to the changing voting pattern and the rise of new post-modern norms and values, new parties entered the field (Ignazi, 1995: 7). On the left side of the political spectrum, the Greens appeared and on the right side, the radical right parties (Ignazi, 1995: 2). News norms, such as freedom and individualism, influenced the voting patterns. Immigration and security became the most valued issues by voters (Rydgren, 2007: 248). The established parties were not able to manage these issues in a proper way, which gave the radical right parties the ability to respond. By offering understandable solutions to unemployment, immigration and tax burden, the radical right parties became an attractive alternative to the established parties (Ignazi, 1995: 8).

Due to individualism, the society became more fragmented (Rydgren, 2007: 248). People have the tendency to divide the society into an “in-group and out-group” (Golder, 2016: 485). The in-group consist of the native population who share similar cultural characteristics and the out-group, on the other hand, consists of the non-native population (Golder, 2016: 485). The division into two groups is emphasized by radical right parties and is used to point fingers. The superior feeling of the in-group relative to the out-group contributed to increased anti-immigrant feelings and support for radical right parties (Golder, 2016: 485).

Another explanation is the modernization of society (Rydgren, 2007: 251). The change from industrial to post-industrial and the appearance of new post-materialist norms caused the segmentation of society into “winners and losers” (Ignazi, 1995: 2). The traditional cultural features are replaced by modern thoughts such as multiculturalism. Those who were disadvantaged by this modernization through a reduced living standard feel excluded and desire the return to a more traditional society (Golder, 2016: 483). These people are frustrated and the established political parties do not offer any solution (Golder, 2016: 483). Radical right parties use this displeasure by emphasising the loss of national identity and the return to traditional values and nationalism (Golder, 2016: 251). Those who feel threatened by this new society will use their vote to make a statement (Rydgren, 1995: 248).

(17)

17 The division between an in and out-group is used to blame the non-native population for economic downfalls. Due to the limited available resources such as jobs and houses, the native population competes with the non-native population (Golder, 2016: 484). The in-group blames the out-groups for the economic problems and will vote for radical right parties if they do not want to compete with immigrants for these limited resources (Rydgren, 1995: 250). Increased competition causes an anti-immigrant sentiment which results in increased support for radical right parties.

The rise of radical right parties is also linked to the decline of the belief in the political system. Ignazi refers to this as the “crisis of confidence” (Ignazi, 1995: 8). Voters lost their confidence in the political system and became pessimistic about democracy (Rydgren, 2007: 251). Voters no longer believed the established parties and the demand for new parties increased (Ignazi, 1995: 8).

The supply explanations of the emergence of the radical right parties focus’ on the party systems and the party itself. As mentioned in the previous paragraphs, the electoral system and electoral rules in a country influence the emergence of new parties. Radical right parties profit from a proportional representation system without barriers, such as an electoral threshold (Rydgren, 2007: 254). The established parties in the field also influence the prospects for radical right parties. If the established parties work closely together and these parties collude, the possibilities for radical right parties enlarge (Rydgren, 2007: 253). Due to their convergence, the differences between those parties fade away. Especially if the established right parties move towards the center of the political spectrum, the chances of electoral success for radical right parties ameliorate (Golder, 2016: 487). Voters turn to radical right parties because of their deviation and by voting on radical right parties, voters express their dissatisfaction concerning the current political process (Rydgren, 2007: 254). A second aspect of the system refers to the media. The media is an important tool for radical right parties to disseminate their views (Golder, 2016: 488). Due to the mass media, radical right parties have the possibility to reach a wider audience, which was not possible before. Due to the commercial purpose of most media, extreme issues of politics are used to attract as much as readers as possible (Golder, 2016: 488). The personalisation of the media with a focus on political leaders also benefits the radical right parties because of the central role of the leader in their party (Rydgren, 2007: 255).

Not only the party system influences the existence of radical right parties, also their own party organization impacts their survival (Rydgren, 2007: 256). Radical right parties have the possibility to emerge but without a proper organization, the party will not persist (Golder, 2016: 489). A network of volunteers, the support of party members and a professional management are all crucial to withstand, especially when the party will govern (Rydgren, 2007: 256).

The final explanation is the political cleavage structure in a country (Golder, 2016: 488). If the chasm in society is stable, the opportunity for new parties to emerge are limited, but if there are new cleavages or the existing cleavages vary, new parties have the possibility to collect many votes (Golder, 2016: 488). As mentioned before, the transition from an industrial

(18)

18 and post-industrial society, globalization and the rise of new issues such as immigration changed the existing political cleavage structure (Ignazi, 1995: 2). Radical right parties benefit from this change through changing voting patterns.

2.5: Hypotheses

Based on the review of the existing literature, the following hypotheses can be formulated to explain the emergence of radical right parties and the effect of electoral systems on the electoral strength of radical right parties.

Hypothesis 1: The number of votes of radical right parties differs between the proportional representation systems and the majoritarian systems.

Explanation: According to Duverger´s law, voting is based on a cost-benefit analysis, whereby the parties with chances of governing are supported. Voting for the radical right parties in majoritarian systems is not rational because of their limited probability of obtaining a majority of the votes and seats. The number of votes for radical right parties are therefore lower in majoritarian systems than for proportional representation as there is no need to shift to stronger parties in the latter system due to the equal distribution of votes.

Hypothesis 2: The relationship between the proportionality of electoral systems and the electoral strength of radical right parties will be positive.

Explanation: Due to the equal distribution of votes in a proportional representation, these electoral systems are characterised by a higher level of proportionality compared to the majoritarian systems. Radical right parties have the best opportunity to achieve seats in a proportional representation and the number of seats will therefore be higher. A positive relationship between the proportionality of the electoral system and the electoral strength of radical right parties is therefore expected. The more proportional an electoral system, the more seats radical right parties are expected to obtain. Majoritarian systems, on the other hand, have a high threshold, which results in a two-party system. Due to the majority rule, radical right parties have limited chances of obtaining seats. The number of seats for radical right parties will, therefore, be lower in majoritarian systems.

(19)

19

Chapter 3: Research design

The following chapter will discuss the methodology and data in order to test the established hypotheses.

The two mentioned hypotheses focus on the electoral strength of radical right parties in the two electoral systems. They intent to hypothesize the relationship between the number of votes and seats between both electoral systems with the aim to determine the influence of proportionality on the electoral strength of radical right parties.

To test these hypotheses, a nested analysis will be applied. A nested analysis consists of a statistical analysis of a large number of cases and a more intensive investigation of a couple of cases (Lieberman, 2005: 435). The level of analysis can, therefore, differ within one research (Lieberman, 2005: 440). Different techniques are used in both analyses in order to find the right information to understand the case. Combining the two approaches can make a positive contribution to the research. According to Lieberman, disadvantages of both approaches can be minimized and the quality of research methods and findings improved (Lieberman, 2005: 436). Using the case studies besides a statistical analysis might be beneficial if the results of the statistical analysis are not robust enough to answer the research question. The Small N research approach gives the researcher more contextually based evidence which improves the confidence of the resemblance of theory and reality (Lieberman, 2005: 442). Besides that, outliers in the statistical analysis can be further examined in order to increase understanding and to generate insights from these outliers (Lieberman, 2005: 435). A common disadvantage of a case study research is the lack of external validity. Practicing a statistical analysis next to a case study can, therefore, improve external validity and strengthen the overall conclusion (Lieberman, 2005: 441). The statistical analysis will also serve as a guide for the case selection, which will result in more focused case studies (Lieberman, 2005: 435).

The nested analysis is used to improve the knowledge concerning the influence of electoral systems on the electoral strength of radical right parties. Based on a Large N analysis, the data would be limited to answer the research question. Adding the case studies will give a more context-based analysis and variation over time which improves the results concerning the influence of electoral systems on the emergence of radical right parties in Western European countries. The case studies will give an elaborated report on the political developments in a country with a focus on the radical right party and the response of mainstream parties. Due to the combination of the two approaches in a nested analysis, it is possible to answer general questions, such as what is the relationship between the number of votes and seats for radical right parties in both electoral systems? And on the other hand more specific questions, what year was the breakthrough of the FPÖ in Austria?

The research will consist of two parts. The first part is the statistical analysis and the second part consists of the four case studies, whereby the first part serves as a guide for selecting the case studies. The following part will discuss the methodology and data by approach.

3.1: Statistical analysis To determine the relationship between electoral systems and the electoral strength of radial

(20)

20 right parties, a panel dataset is used. The dataset includes 15 countries, all located in western part of Europe and covers the period 1980 through 2015. The dataset includes 540 country-year observations. The data are country-year-end values. The data is collected from The Comparative Political Data Set (CPDS). The analyses based on the dataset are presented in the tables in the Appendix.

The cases selected for this research are Western European countries. The selection of the countries is based on their geographical position. The countries included in the dataset are Austria, Belgium, Denmark, France, Germany, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland and the United Kingdom.

Descriptive statistics:

To measure the difference between the number of votes and seats for radical right parties in the electoral systems, the variables of CPDS dataset are used. Table 8.1 gives an overview of the variables and descriptions.

Dependent variable

The total number of votes or the total number of seats for radical right parties are used as the dependent variable.

The selection of radical right parties is based on the classification of the authors. The authors make a classification of political parties in 20 categories: “Social democratic, Left-socialist, Communist, Post-communist, Agrarian, Conservative, Religious, Liberal, Protest, Green, Ethnic, Right populist, Regionalist, Feminist, Monarchist, Personalist, Pensioners, Non-label, Electoral alliance and others” (Armingeon, et al., 2017: 49). The Right populist, called right in the research sample, is the category to which the radical right parties belong. The totalright_votes is the sum of all votes for all right parties in a country. The totalright_seats is the sum of all seats for all right parties in a country. These two variables are used to determine the difference in the number of votes and seats of radical right parties. Table 8.3 gives an overview of the total votes and total seats for right parties per country.

A remark has to be made concerning the classification of the United Kingdom by the CPDS dataset. UKIP was classified as a protest party, but this thesis classifies the party a radical right party. Therefore, it belongs to the right category and will be included in the study. The argumentation regarding the classification of UKIP as a radical right party is further expressed in the case study about the UK.

The prop variable indicates the electoral system of a country. This is divided into two categories. If coded 0 and 1, the country has a majoritarian system. If coded 2, the country has a proportional representation. In 2015, there were two countries with a majoritarian system and thirteen countries with a proportional representation. Italy is the only country who has changed the electoral system between the period 1980 and 2015. In 1994 the country implemented a majoritarian system but this has been reversed in 2005.

Independent variable

(21)

21 Disproportionality means “the deviation of the parties’ seat shares from their votes shares” (Aikaterini, 1999: 9). Perfect proportionality indicates a system where the share of votes is transferred into the same share of seats (Aikaterini, 1999: 9). As mentioned in the literature review, indicating the proportionality of an electoral system can be based on different measures. Two different measures are used to indicate the proportionality of both systems. Both variables originate from the CPDS dataset.

The first variable is the effective number of parties on the seats level. This variable indicates the number of equal sized parties in an electoral system based on the calculations of Laakso and Taagepera. A higher number of the effective number of parties indicates a higher level of proportionality of the electoral system. In the research sample, this variable takes the value of 2.091 to 9.08.

The second variable is the index of absolute disproportionality. This index indicates the proportionality of electoral systems which is based on the effective number of parties on the votes level minus the effective number of parties on the seats level. The variable has to be indicated as follows “a score of 1 means that on the seats level there is, on average, one (effective) party less than on the votes level” (Armingeon, et al., 2017: 9). The allocation of seats is in this case not based on the equal distribution of shares of votes into shares of seats. A higher value of this variable indicates a higher level of disproportionality and less fragmentation. This index of disproportionality differs between countries. In 2015, the index of France was 2.45 and the index of the Netherlands 0.25. This indicates that the Dutch system is more proportional than the French system.

Table 8.2 presents the summary statistics. For each variable, it presents the number of observations, the mean, standard deviation and the minimum and maximum.

The statistical analysis will expose the differences between the proportional representation and the majoritarian system and gives an indication in which system radical right parties have better circumstances to flourish.

3.2: Case studies

Besides the statistical analysis, multiple case studies are selected to test the hypotheses. The case studies are used as an extra manner to test the emergence of radical right parties which would not be uncovered by the statistical analysis. The use of case studies increases variation, more context and evolution over time which will improve the knowledge about the influence of electoral systems on the electoral strength of radical right parties.

The cases are selected based on the strategy of Lieberman. He makes a distinction between a Model-testing Small N analysis and a Model-building Small N analysis (Lieberman, 2005: 437). The Model-testing analysis is used when the researcher is satisfied with the model and the results of the statistical analysis. The results are then significant and the researcher is already certain about the fitness of their model (Lieberman, 2005: 442). Adding the Small N is a manner to further tests the robustness and to increase contextually based evidence (Lieberman, 2005: 442). By using both approaches, the researcher can be certain their theory is statistically correct and is corresponding with the reality. The cases should be selected

(22)

22 based on their accuracy of fit (Lieberman, 2005: 444). Countries that are on or close to the line, which indicates the relationship between the interested variables, should be chosen for further Small N analysis (Lieberman, 2005: 444). The researcher is not interested in case off the line, these can be seen as noise (Lieberman, 2005: 444). Investigating the on the line cases gives the possibility to check the robustness and to fine-tune the theory if necessary.

The Model-building Small N strategy is used if the results of the statistical analysis are not satisfying. The established theory is weak and the hypotheses cannot be confirmed based on the Large N analysis (Lieberman, 2005: 443). An in-depth research will increase the knowledge concerning the theory and based on the case studies the theory can be confirmed or adjusted. If the case studies generate the same results, the researcher can be certain their theory works and corresponds with the context. If the results do not correspond to each other, the researcher should concern testing a new model with a Large N analysis (Lieberman, 2005: 437). Countries should be selected, based on variation and comparison. Countries which are on and off the line should be selected (Lieberman, 2005: 437). The Small N analysis can explain these cases which were not fully explained by the Large N analysis. The comparison between on and off the line cases gives the researcher a complete picture and extends the robustness of the general conclusions (Lieberman, 2005: 445). According to Lieberman, the cases should be selected based on the dependent variable because of the interest in differences in outcome (Lieberman, 2005: 445).

The statistical analysis gives an indication of the relationship between the influence of proportionality and the electoral strength of radical right parties. But more research is needed to increase the robustness of the results. Therefore the Model-building Small N strategy is applicable in this case. Adding the case studies will increase our theory concerning the emergence of radical right parties and the influence of the electoral system. Combining the two approaches will, in the end, benefit the conclusion.

As described above, the case selection of the four case studies is based on the statistical analysis. The relation between the proportionality of the electoral system and the number of seats for radical right parties is determining. The red line in the following figure illustrates the negative relationship between the proportionality and the electoral strength of radical right parties.

(23)

23 Figure 3.1: The relation between the total seats of radical right parties and the index of absolute disproportionality, per country

Four case studies are selected, which were on and off the line and with different electoral systems to determine the electoral strength of radical right parties and the effect of electoral systems. The cases selected are France, the United Kingdom, the Netherlands and Austria. France and the Netherlands both fit on the line and the United Kingdom and Austria are more off the line. The variation in electoral systems will benefit the research and results.

To test the hypotheses, different documents and available materials will be used. Such as election results, voting research and established literature.

(24)

24

Chapter 4: Results

The following chapter will present the results of this research. The results will be divided in the results of the statistical analysis and the results of the case studies.

4.1: Statistical analysis Based on the CPDS dataset, the following results will be presented concerning the emergence of radical right parties and the influence of electoral systems on the electoral strength of radical right parties. First, the difference in the number of votes and seat for right parties in the electoral systems will be discussed and afterwards the relationship between proportionality and electoral strength in Parliament.

There is an increase in the number of votes for right parties, which counts for almost all countries. Figure 4.1 shows the trend of right votes from 1980 up to and including 2015. All researched countries are included. In 1980, the radical right parties received 0,493 percent of the total votes. In 1998 this was already 6,81 percent and this number increased to 10,17 percent in 2015. Potential reasons for this increase are discussed in chapter 2 and will further be discussed in the case studies. Although almost every country has experienced an increase in the number of votes for radical right parties, the average number of votes still differs. Table 8.3 gives an overview of the average percentage of votes and seats for right parties per country. The percentage of votes differs greatly from country to country. In Norway, the radical right parties together received an average of 1,36 percent of the total votes, as in Italy these parties received an average of 12,24 percent of the total votes.

Figure 4.1: Trend total votes of radical right parties between 1980-2015

Duverger’s law could explain the reason behind the difference in the number of votes between countries. As hypothesized, the votes for radical right parties will be lower in a majoritarian system because of rationality. Voters prefer parties who have the possibility to govern, this influences their voting strategy. Radical right parties have a limited chance of governing in a

(25)

25 majoritarian system and people will, therefore, give their votes to other parties. This is not necessary in a proportional representation because of the equal chance of governing for all parties.

The following table presents the number of votes and seats for radical right parties by electoral system.

Table 4.1: Total right seats and total rights votes by electoral system

Prop Totalright_seats Totalright_votes

Modified proport 3.736 7.365

Proportional rep 4.558 4.722

Total 4.430 5.133

Based on the dataset of the 15 countries, it suggests that the difference in the number of votes between countries cannot be explained based on the electoral systems. Table 4.1 shows that the radical right parties received an average of 7.4 percent of votes in a majoritarian system. In countries with a proportional representation, the radical right parties received an average of 4.7 percent of the total votes.

The following table presents the difference in sum of radical right votes per electoral system. Table 4.2: The difference in total right votes by electoral system

Majoritarian system Proportional representation Difference Totalright_votes 7.365 4.722 2.644***

This table reports the difference between selected variables. *** Statistical significance at 1% level. ** Statistical significance at 5% level. * Statistical significance at 10% level.

To explore whether or not there is a significant difference in the sum of radical right votes between countries with a majoritarian system or proportional representation this paper uses a Univariate Test. On average there is a 2.6 percentage point difference in the sum of radical right votes by the electoral system. The average percentage of radical right votes is for a majoritarian system 7.4 percent, which is statistically significant higher compared to the proportional representation system. Therefore, I reject the Duverger’s law at a 1 percent significance level. Based on the statistical analysis, the opposite counts as the number of votes for radical right parties are higher in countries with a majoritarian system than in countries with a proportional representation.

(26)

26 Table 4.3: Correlation between the total votes and total seats for radical right parties

Totalright_seats Totalright_votes

Totalright_seats 1,0000

Totalright_votes 0,9340*** 1,0000

This table reports the correlation matrix of selected variables. *** Statistical significance at 1% level. ** Statistical significance at 5% level. * Statistical significance at 10% level.

Table 4.3 indicates the correlation between the number of votes and number seats of radical right parties. There is a significant positive correlation of 0,934. This indicates that an increase in the number of votes for a radical right party is accompanied by an increase in the number of seats for that party. In order to estimate the relation between both variables, a Prais Winston regression is used. This regression estimates the parameters and addresses autocorrelation. Table 4.4 illustrates the results of this regression. There is a significant positive relationship, which indicates that increased number of votes is accompanied by an increased number of seats.

Table 4.4: Regressions on the relation between total votes and total seats of radical right parties Totalright_seats dependent variable Totalright_Votes 0.92*** (38.83) Constant -0.45 (-0.74) 𝑅2 0.7327 Number of obs. 540

This table presents Prais Winston results. The dependent variable is total right seats. *** Statistical significance at 1% level. ** Statistical significance at 5% level. * Statistical significance at 10% level. T-statistics are in parenthesis.

The main question is if this positive relation between the number of votes and seats counts for both electoral systems. The following part will zoom in on the difference in the number of seats between a majoritarian system and a proportional representation. The statistical analysis reveals a strong difference in the number of seats between both systems. Radical right parties in a majoritarian system have on average obtained 3.7 percent of the total seats, where the radical right parties in a proportional representation have received 4.6 percent of the total seats. It seems that radical right parties have the best possibility to flourish in proportional representation systems as all votes are converted into seats. The majoritarian system is characterised by a discrepancy between the number of votes and seats for radical right parties.

(27)

27 Obtaining seats in this electoral system is more difficult as relatively more votes are needed to obtain seats compared to a proportional representation. The discrepancy between the number of votes and seats by the electoral system is displayed in table 4.1.

To estimate the influence of electoral systems on the electoral strength of radical right parties, the relationship between proportionality and seats is estimated. As explained in the Research Design, there are two variables used to indicate the proportionality of an electoral system, the effective number of parties on the seats level and the index of absolute disproportionality. Table 4.5 presents the relationship between the number of seats of radical right parties and the index of absolute disproportionality.

Table 4.5: Regressions on the relation between the total seats of radical right parties and index of absolute disproportionality

Totalright_seats dependent variable Dis_abso -1.36*** (-2.66) Constant 5.34 (2.43)** 𝑅2 0.0016 Number of obs. 540

This table presents Prais Winston results. The dependent variable is total right seats. *** Statistical significance at 1% level. ** Statistical significance at 5% level. * Statistical significance at 10% level. T-statistics are in parenthesis.

There is a significant negative relationship. This indicates that an increase of disproportionality results in a decrease of the number of seats of radical right parties. The disproportionality of the system shows that there are fewer parties on the seats level than on the votes level. Countries with a higher index of absolute disproportionality, which are the countries with a majoritarian system, are characterized by a lower number of seats for radical right parties. Countries with a low index of absolute disproportionality are characterized with a higher number of seats for radical right parties.

The relationship between the effective number of parties and the number of seats of radical right parties shows a similar relation. The effective number of parties indicates the proportionality of an electoral system. A higher number indicates a higher level of proportionality. Countries with a proportional representation are featured by a higher number of effective parties.

The emergence of radical right parties in European countries and the influence of electoral systems on this process