POPULISM AS A CONSTANT COMPANION OF DEMOCRACY

Transcription

1 POPULISM AS A CONSTANT COMPANION OF DEMOCRACY Empirical evidence from Switzerland between 1947 and 2011 Paper prepared for the ECPR General Conference, Montreal, August 26-29, 2015 Edward Weber NCCR Democracy, University of Zurich weber@nccr- democracy.uzh.ch Abstract This article analysis the development of the degree of populism in the Swiss political discourse between 1947 and Based on a quantitative content analysis of the election manifestos of all relevant Swiss parties in this period, the results show that it seems exaggerated to speak of a current populist Zeitgeist as it has been argued. The new radical Right in Switzerland, which dominated the populist discourse during the last two decades is just the newest version of highly populist new out- sider- parties from the whole political spectrum which entered the Swiss party system since the after war period. Populism therefore has to be seen rather as a constant companion of democracy then a (con- ) temporary phenomenon. 1

2 Introduction Simultaneously with the rise of the populist radical Right in Western Europe, the phenomenon of pop- ulism got a fast increasing attention in political science. Not surprisingly, the resulting empirical litera- ture focused initially on the political discourse of this new party family. Some political scientists even went so far as to inductively incorporate nationalist, conservative or exclusionist ideological aspects of the new radical Right into the definition of populism (e.g. Albertazzi and McDonnell 2008, Taguieff 1995). Since some years however, political scientists achieved a broadly accepted consensus that pop- ulism is neither left nor right per se. Mudde s (2004: 543) definition of Populism as a [thin] ideology, which can be combined as well with socialism as with nationalism or other ideologies, is now the ref- erence point for most of the populist literature (Kriesi 2014, Decker 2006, Rooduijn 2013, March 2012). In line with this definition, the empirical literature expanded their object of study and analysed also the degree of populism in the discourse of contemporary left and moderate parties (March and Mudde 2005, March 2007, 2012, Rooduijn et al. 2012, Pauwels 2014). What they all have in common however is that they focus exclusively on the most recent past and their time horizon does not exceed the last two decades. We do not know thus whether, with the recent emergence of the new populist parties, the general political discourse reached an unprecedented level of populism compared to pre- vious decades and if we therefore live in a populist zeitgeist as Mudde (2004) claimed 1. Or, if rather the members of the new party family of the radical Right are just as populist as other parties have been at the time when they were still at the periphery of the party systems and not yet integrated in the power structure of their states, and if we are therefore now just dealing with the newest version of populism as a constant shadow of democracy (Canovan 1999). In order to shed some light on this question and get a broader understanding of the general status of populism in politics and its devel- opment over time, I will analyse the populist discourse in the election manifestos of the Swiss parties from the post- war period until the present. The article is structured as follows. First, I will discuss the definition and conceptualization of populism. Second, I will outline the theoretical explanations for three hypotheses regarding the devel- opment of populism over time. Then, I will proceed with a discussion of the design and method of the study. I will particularly elaborate on the characteristic of the Swiss party system, on the elections manifestos as the data basis and on the measurement and calculation of the degree of populism as the dependent variable. Fourth, I will describe the results of the longitudinal content analysis and dis- cuss the development of populism in Switzerland over time. Finally, I will summarize these findings and discuss the limitations of the study. Definition of populism As the basis of the measurement of populism, I will use Mudde s (2004: 543) definition of Populism as "a [thin] ideology that considers society to be ultimately separated into two homogenous and antagonistic groups, 'the pure people' versus 'the corrupt elite', and which argues that politics 2

3 should be an expression of the volonté générale (general will) of the people 2. In recent years, this definition has established itself as reference point of the majority of the political science literature about populism (e.g. Kriesi 2014, March 2012, Pauwels 2011, Rooduijn et al. 2012). Based on this definition, populism can be disaggregated into two elements: First, the antagonistic relationship between the people and the elites and second, the idea of popular sovereignty (cf. Kriesi 2014: 362) 3. Populism therefore means firstly to oppose the interests of the people as a homogenous actor with a common will - to the interests of the elites. On the basis of this conflict of interest the elite is criticized for not caring about the people, to have no idea about the needs of normal people or even of actively betraying it. The antagonistic relationship between the elite and the people as the first element of populism has therefore a lot in common with anti- elitism. If criticizing elites is populist or not, crucially depends however on the question if the critique contains any reference to their negative relationship with the people. What is meant by the concepts of the people and the elite varies. Depending of the thick ideology of the speaker (see below), the notion of the people can for example refer to the working class, the nation or an ethnic group (Rooduijn et al. 2012). To the elite may poten- tially belong all those groups which have a privileged and powerful position compared to the rest of the society. This can be the cultural elite as well as the economic elite. But in established democracies it is clearly most often the political elite, who is attacked in a populist manner, because of the fact that even in semi- direct democracies like Switzerland - they are deciding most of the decision without consulting the people. This leads us to the second basic element of populism. The logical consequence of the Manichean populist worldview concerning the antagonistic rela- tionship between the good people and the bad elite is the claim for unrestricted popular sovereignty. The citizens have been shut out of power by corrupt politicians und an unrepresentative elite [...] (Canovan 2002: 27). Thus whereas populism criticizes all elements of liberal democracy (Mudde 2004), it supports more direct, transparent and simple connections between the people and the political power. In the populist worldview, the balance of power in the state should be changed in favour of the people and to the disadvantage of political, economical and other societal elites as well as interna- tional organizations and their laws. Institutional reforms for an introduction or expansion of direct democracy can be one possible way to attain this goal (Canovan 1999: 2). But there are also more authoritarian options to further people s power, as installing a political leader who acts as plebiscitary ruler and spokesperson of the volonté general (March 2012: 8). This is why populism is inherently neither necessarily democratic nor undemocratic (cf. Kaltwasser 2011: 194). Not only can populism be related to democratic as well as undemocratic visions of the popular sovereignty, but it is also not per se bound to left or right ideology. In the theoretical literature about populism, it is sometimes argued, that populism does not only exclude the elite from the people but also a second group of "outsiders" like immigrants and religious or ethnic minorities, and that this exclusion is also a defining element of populism (Jagers and Walgrave 2007, Albertazzi and McDonnell 2008, Taguieff 1995). Since this would a priori prevent all left and moderate parties from being popu- 3

4 list, most scientist nowadays agree by contrast, that exclusionist ideology is only part of right wing populism but not of the core of populism (Kriesi 2014, March 2012, Mudde 2007, Rooduijn et al. 2012, Taggart 2000). I thus follow Mudde s definition of populism as a thin ideology, which means that populism is per se neither left nor right and can be combined with other ideologies as socialism, liber- alism or nationalism and thereby become a thick ideology (Mudde 2004). Development of populism over time A) Denationalization and the new populist area Today s populism in Western Europe is generally seen as closely related to and induced by the process of globalization. First, the increasing importance of the European Union and other supranational insti- tutions led to a reduction of the manoeuvring space of national governments (Kriesi 2014: 365). This effect is enhanced by global capitalism, which further restricts the sovereignty of democratic nation states. Consequentially, denationalization led to a decline of the parties accountability to their na- tional constituency of voters. Since the scope for action of national governments is strongly restricted in the globalized world, the parties are no longer able to pursue political goals of their electorates (ib.). This decrease in the representation function of the parties decisively affected the relationship be- tween the people and their political elites in western democracies: With the transfer of decision- making competences to the supranational level, very serious problems are posed with respect to the congruence between the government and the people, the inclusion and participation of national par- liaments and national electorates in the political process, the transparency of the decision- making process, the accountability of national executives and of non- representative supranational executive bodies to national parliaments and national electorates, and the multiple executives responsiveness to the citizens needs (Kriesi 2013: 212). The citizens got the impression that the established parties are all alike, not responsive to their demands and that they all betray the public behind the scenes (Kriesi 2014: 367). With the increase in voter s dissatisfaction with politics, it became much easier in the last years for outsider actors to protest in a populist manner against the increasingly similar and unresponsive established parties (Mair 2002: 88). As a consequence, we witnessed the rise of the new populist right party family during the last two decades. The subsequent electoral growth of these parties with their strong populist discourse further provoked an increased populist rhetoric of the established parties. In an attempt to counter the populist challengers, mainstream parties resort more and more to populism in their political discourse (cf. Mény and Surel 2002). Following this logic, we live today in a populist Zeitgeist in Western Europe, as Mudde (2004) stated. Globalization therefore leads to a reduction of the representative function of the established par- ties, which then again prepares the breeding ground for successful populist actors. This in turn forces also the other parties to adopt a more populist rhetoric. In regard to the development of populism over time this leads us to the following hypothesis: 4

5 Hypothesis 1: The political discourse of the parties became more populist in the recent past. B) Populism as a constant companion of democracy In contrast to the above stated line of argument, populism can also be seen as a more time- independent phenomenon that constantly accompanies democracies like a shadow (Canovan 1999: 16). According to this view, real popular sovereignty in Western European countries did not only get lost in the process of denationalization, but hardly ever existed before. Several permanent limitations of people s power in existing formal democracies constituted a perpetual breeding ground for a popu- list discourse already before globalization took its full effect in the 1990s. First, strong inequalities in the economic sphere restricted the egalitarian promise of democracies since their beginning. The eco- nomic elites were always able to transform their economic power into political power and thus reduc- ing the political influence of the rest of the citizens. Second, popular sovereignty is further restricted by the liberal aspects of our Western European democracies. As Mudde (2004: 561) noted: Despite all democratic rhetoric, liberal democracy is a complex compromise of popular democracy and liberal elitism, which is therefore only partly democratic. Especially the element of political representation with its inherently oligarchic dimension (Papadopoulos 2002: 48) is opposed to the democratic promise of unrestricted popular sovereignty. Above that, the constitutional protection of minority rights and the independence of important institutions like courts or the central bank from the popular will, are all reducing the sphere of politics, which in democracies means a loss of popular sovereignty. And finally, national politics was even before globalization not unrelated to international develop- ments. Global economic crisis and the decisions of other countries always restricted national - and therewith also popular - sovereignty in Western European democracies (Dahl 1994: 24). Since popular sovereignty was always limited by social inequality, the liberal pillar of Western Eu- ropean democracies and international developments, there has also been a constant breeding ground for populist attacks against the established order in the name of the people. The gap between ha- loed democracy and the grubby business of politics (Canovan 1999: 11) thus potentially leads to the formation of new parties, promising in place of the dirty world of party manoeuvring the shining ideal of democracy renewed (ib.). These new challengers in the party system are able to attack the estab- lishment for their depolitization of important conflicts and demand institutional reforms that should bring politics closer to the people (cf. Kriesi 2014: 368f). Seen in this light, democracy is a self- correction system in which new outsider parties regularly enter the party system criticizing the exist- ing democratic limitations to enhance popular sovereignty (Canovan 1999: 14). The strong populist discourse of the new radical right party family would therefore not be an exception but a general characteristic of new parties, which just entered the political system. Based on these thoughts, I will test the following hypothesis: Hypothesis 2: New parties entering the political system are more populist in their political dis- course than the other parties. 5

6 Once being part of the political system, these new challenger parties will however hardly hold their strong populist discourse forever. If we analyse the parties for a longer time period, there are some reasons why we should expect a dynamic transformation of populist challengers from the pe- riphery of the party system to established parties at the centre of it. Accordingly, the high level of populism in the discourse of new parties should decreases over time. The first reason for this assump- tion is, that in order to convert electoral gains into governmental power, new parties have to town down their populist rhetoric and thus become an acceptable coalition partner (Rooduijn et al. 2012: 14). Second, once the party is able to achieve governmental status on the local, regional or national level, their new responsibility may further socialize them into regular parties and they therefore be- come even less populist. This transformation of new populist outsider parties into non- populist established actors is not one of clearly distinguishable steps but rather an incremental process. This is because the decision to town down its populist discourse in order to become a possible coalition partner is not taken from one day to another but during a long- term process with often severe internal disputes over the goals and means of the party. Typically, outsider parties normally start their integration process in taking part in local governments first and thereafter successively extend it to higher levels (Kriesi 2014: 371). They herby undergo a process of integration from the periphery into the core of the political system and the degree of populism in their discourse reflects their current stadium in this development. If it is indeed the case, that parties become less populist after entering the political system, I will test with the following hypothesis: Hypothesis 3: The longer a party takes part in the political system, the less populist is its political discourse. Design and method I selected Switzerland as the country for which I will test the above stated hypotheses. This has some advantages regarding the study of the development of populism in the parties discourse over time. First, Switzerland is - due to its numerous political cleavages - the country with the highest number of parties during the period from the end of the Second World War until the end of the 20 th century (Li- jphart 1999: 76). This allows me to compare the degree of populism of many cases in the same coun- try. Second, the Swiss party system is a very stable one with long- term continuity (Caramani 2004: 116). Parties, which once entered the party system on the national level, often tend to stay there for several decades. This is important for testing the integration- hypothesis, since it enables me to ana- lyse the political discourse of the same parties during several elections 4. For testing the three hypotheses about the development of populism over time, I considered the time period from the first election in the after war period in 1947 to the last election in This is the maximal time span during which party documents on the national level are systematically availa- ble. This means that the analysis contains 17 national elections. I selected, with a few exceptions 5, every party that gained at least 1% in two successive elections. With this criteria a total of 14 parties 6

7 are included in the study, which together represent between 90.5% (in the election year 1991) and 95.7% (in the last elections of 2011) of the votes in the national elections during the research period. The considered parties can be shortly summarized as follows: Six of these parties were already founded before the beginning of the research period: The social democratic party (SP), the Christian democrats (CVP), the Free democratic party (FDP), the formerly rural Party of Farmers, Traders and Independents which transformed in the 1990s to the radical right Swiss people s party (SVP), the Ring of Independents (LDU), and the Evangelic party (EVP). SP, CVP, FDP and SVP constituted the core of the Swiss party system during the research period and with only few exceptions 6 formed together the national government during this time. The remaining 8 parties appeared during the period of analysis or immediately before: Three of them are leftist parties: The communist Party of Labour (PDA), the Progressive Organisation of Switzerland (POCH) and the Green Party (GP). Another three are belong- ing to the conservative or radical- right: The National Action/Swiss Democrats (SD), the Federal Demo- cratic Union of Switzerland (EDU) and the Automobile Party/Freedom Party (FP). The moderate parties of the Green- Liberals (GLP) and the Conservative Democrats (BDP) emerged on the political landscape in the last decade. I used a quantitative content analysis of election manifestos to measure how populist a party is at a given time point. Compared to other types of party documents, election manifestos have two basic advantages as a source of measuring parties ideology over time in general and populism of Swiss par- ties in particular. First, it is the party document which is most standardized over time and across par- ties. During the whole research period of more than 6 decades, almost every party - even those that just entered the party system produced similar documents with their general political goals (cf. Rooduijn 2012: 6f). Second, since election manifestos are to a large degree independent of daily party politics, they contain more fundamental (populist) critique of the political system and its establish- ment than specific critique concerning a single party or a specific policy of the government. And final- ly, election manifestos are nearly the only documents in which the deeply regionally fragmented Swiss party organisations take a common position, especially regarding the beginning of the research peri- od 7. These manifestos have been analysed by trained coders, who used a comprehensive codebook. The criteria for the decision whether a statement is populist or not, closely follows the above outlined two aspects of people- elite antagonism and popular sovereignty. A statement is therefore coded as populist, if it either criticizes elites of not behaving in line with the interest of the people or if it claims more power for the people in the state. The antagonistic relationship between the people and the elites - as the first element of populism - was operationalized by the following questions: Do elites pose a threat or burden to the people/the society according to the statement? and Are the elites described in the statement as not belong to/not being close to/not knowing the needs of/not caring about/not speaking on behalf of/not empowering/deceiving the people?. Popular sovereignty was measured by the questions: Does the statement argue that the people should have/gain/not loose power?, Does the speaker give the people the competence to act or decide on a specific political 7

8 issue?, and Does the statement demand institutional reforms for more participation of the people in politics?. In line with the conceptualization of populism as outlined above, the term people in the code- book also includes people as a an ethnos ( The Swiss ), a societal layer like the employees, people as a function ( the voters ), a hypothetical prototype of the people ( the common man ) or any other term which can stand for the majority in the society. The criticised elites may be political ( the es- tablished parties ), financial ( the capitalists ), economic ( the trusts ) or cultural ( the academics ) elites, the media or just unspecified elites ( the mighty ones ). Two additional specifications have to be made concerning the distinction between populist and non- populist statements. Concerning the first aspect of populism it is essential that not only the cri- tique of elites is linked to the interests of the people, but also that the target of the critique is one of the above mentioned elites in general. Therefore neither critique directed towards the whole political establishment without reference to the people, nor critique of a part of an elite - like a specific party - are considered as populist statements. Second, a claim for more popular power is only considered populist if the demand is related to the political sphere. Demands for more power of normal people in the economy like the postulation of a democratic economy are regarded as leftist ideology, comparable to claims against privatizations of industries or an economically more equal society. Only if the speaker links the economic inequality to politics and emphasizes the democratising effects an equal and citizen- owned economy would have, can the statement be defined as populist. The degree of populism in an election manifesto is calculated by the percentage points of para- graphs with contain a populist statement in relation to the total number of paragraphs. Paragraphs as units of measurement have the advantage of being more objectively traceable and thus enhancing the reliability of the measurement (Rooduijn 2013: 7). Furthermore, since most of the populist statements are elaborated in several sentences, smaller measurement units like words or sentences (e.g. Jagers and Walgrave 2007) are less appropriate (cf. March 2012: 15, Rooduijn 2013: 17). The time point when the parties entered the political system - which is crucial for the analysis of the second hypothe- sis - is defined as the election year, when the respective party for the first time gained at least 1% of the votes in the national parliamentary elections. This criterion is equivalent to the one, which a party has to fulfil in order to be considered in this study (see above). In order to control for differences in the length of the election manifestos (measured by the number of their paragraphs), I will follow Rooduijn et al (2012: 8) and assign more weight to populist paragraphs in longer manifestos 8. This seems appropriate since more detailed programmes tend to contain more specific policy goals and therefore a lower share of populist statements (March 2012: 21). The thus calculated populism- scale theoretically ranges from 0 to 100. Empirically, roughly 40% of totally 132 election manifestos 9 contain no populism at all, while only the four most populist manifes- tos have values of 30 or more: Two of them from the PDA (1963: 45, 1975: 35), one of the POCH (1975: 32) and the latest manifesto of the SVP (2011: 30) The overall degree of populism at a given point in time as the dependent variable of the first hy- 8

9 pothesis is calculated by the average level of populism in the manifestos of all considered parties. However, it might be argued, that for the analysis of the development of populism over time, parties with higher levels of electoral support may influence the political discourse in a country to a higher degree than smaller ones. Following this logic, not only the level of populism of each party would be important, but also the question how successful strongly populist parties are at the polls. Therefore, the degrees of populism at the 17 election years are also calculated as a weighted average, based on the electoral support for the parties in the respective election. Results Figure 1 shows the values of populism at each of the considered 17 election years, calculated by the average of the manifestos of all parties, which have gained a minimum of 1% of the votes. Looking at the development of these values over time, we cannot recognize a clear trend during the investigation period of During the whole period, the populist discourse never increased or decreased over more than three election years. 12"" Average"degree"of"populism" 10"" 8"" 6"" 4"" 2"" 0"" 1947" 1951" 1955" 1959" 1963" 1967" 1971" 1975" 1979" 1983" 1987" 1991" 1995" 1999" 2003" 2007" 2011" Elec=on"Year" Figure 1. The unweighted average degrees of populism in the election manifestos of all parties. The dotted line represents the average value of the whole research period (4.47). From 1947 until the 1970s the degree of populism, which was during this time dominated by the social democrats (SP), the communists (PDA) and to a lesser degree by the moderate LDU developed un- steadily. The highest average value of populism since the after war period was reach in 1975, which is due to the fact, that in this election year the new radical left Swiss Progressive Organization (POCH) emerged on the political landscape with a distinctively populist election manifesto. Also because the POCH toned down its populist discourse in the following elections, the average level of populism shrunk until From then on, populism again became a little bit more important in the Swiss politi- cal discourse. This is above all due to the emergence of the Freedom Party (FP) as another strongly populist party. At the end of the research period the populist discourse was mainly dominated by the Swiss people s party (SVP), which transformed during this time from a moderate rural party to a typi- cal representative of the new populist radical right party family in Western Europe. 9

10 For testing the first hypothesis, we have to focus more specifically on the degree of populism in the last years compared to the rest of the research period. Regarding the results in figure 1, there is no sign of a populist Zeitgeist. In none of the election years of the 1990s and of the 2000s do the manifes- tos of the parties contain a substantially higher proportion of populist paragraphs compared to the rest of the research period. The highest values in the last two decades are those of 2003 and 2011 and even they are - with values of 4.5 and 4.9 respectively - not much above the average value of the whole research period (4.47). This finding corresponds to those of the study by Rooduijn et al. (2012). In their analysis of election manifestos, which is restricted to the last two decades, but includes five countries, they could not find an increasing trend towards more populism in the last years either. While we are not able to find an continuous trend for the unweighted average degree of popu- lism during the research period, a distinct change in the ideological character of the parties dominat- ing the populist discourse becomes visible in figure 2 which differentiates between left (PDA, SP, POCH, GP), radical right (SD, FP, EDU, SVP) and moderate parties (FDP, CVP, LDU, EVP, GLP, BDP). In the after war period the communists and social democrats (PDA and SP) were by far the most populist parties, before they became clearly less populist at the beginning of the 1970s. Conversely, the radical right parties emerging at this point in time, became ever more populist in the last four decades and are nowadays on average clearly more populist than leftist parties. During the whole research period the moderate parties remained limitedly populist. 25"" Average"degree"of"populism! 20"" 15"" 10"" 5"" Moderate"Par=es" LeD"wing"par=es" Right"wing"par=es" 0"" 1947" 1951" 1955" 1959" 1963" 1967" 1971" 1975" 1979" 1983" 1987" 1991" 1995" 1999" 2003" 2007" 2011" Elec=on"Year" Figure 2. The unweighted average degree of populism in the election manifestos of left (PDA, SP, POCH, GP), radical right (SD, FP, EDU, SVP) and moderate (FDP, CVP, LDU, EVP, GLP, BDP) parties. A clear time trend is also visible on the overall level of populism if we take into account the electoral support of the parties. When the populism scores of the manifestos are weighted by the vote shares of the respective party, the average degree of populism - in line with the first hypothesis - increased by tendency during the last two decades (see figure 3). This development does however mainly reflect the electoral growth of the radical right Swiss People s Party (SVP) as well as the contemporaneous increasing level of its populist discourse. If we exclude this party from our analysis, the trend towards more populism in the election manifestos vanishes (see figure 4). Indeed if we look at the other par- 10

11 ties, the Social Democrats (SP) are the only one with an increasing degree of populism since the 1990s (see figures 5 and 6), while still being far from surpassing the degree it reached in the after war period. 12"" Average"degree"of"populism" 10"" 8"" 6"" 4"" 2"" 0"" 1947" 1951" 1955" 1959" 1963" 1967" 1971" 1975" 1979" 1983" 1987" 1991" 1995" 1999" 2003" 2007" 2011" Elec=on"Year" Figure 3. The average degree of populism of all parties, weighted by their vote shares in the respective election. The dotted line represents the average value of the whole research period (3.84). 12"" Average"degree"of"populism" 10"" 8"" 6"" 4"" 2"" 0"" 1947" 1951" 1955" 1959" 1963" 1967" 1971" 1975" 1979" 1983" 1987" 1991" 1995" 1999" 2003" 2007" 2011" Elec=on"Year" Figure 4. The average degree of populism of all parties except the radical right Swiss People s Party (SVP), weighted by their vote shares in the respective election. The dotted line represents the average value of these parties over the whole research period (3.05). If we shift our focus to the parties which were founded after the beginning of the research period, we can first of all state the following: At the time when they entered the political system they are on av- erage clearly more populist than the older parties. The election manifestos of the 8 new parties at the time when they first gained a minimum of 1% of the votes, had on average a populism score of 8.5, while the average value of the old parties during the whole period since 1947 is 2.7. However, if we look at these 8 parties separately (see figure 5), we can categorize them in three groups with large differences. The populist scores of the PDA (1947: 13), the POCH (1979: 32) and the FP (1987: 15) are many times above the all- time average of the established parties. SD (1971: 3.4) and GP (1983: 3.0) are still but only slightly above the reference value, while EDU (1987: 1), GLP (2007: 0, not in the fig- 11

12 ure) and BDP (2011: 0, not in the figure) were (nearly) completely non- populist parties at the time, when they joined the party system of Switzerland. Overall, the second hypothesis can thus be general- ly confirmed, although it does not hold true for all of the new parties. Bearing in mind the argumentation for the third hypothesis, the absence of populism in the mani- festos of GLP and BDP may be explained by the fact, that they were both founded as splits of an estab- lished party and were therefore already since their very beginning represented in the national parlia- ment, or - in the case of the BDP even in the federal government. These two parties did not have to go the way from the periphery of the political system to its centre. Instead they started being part of the establishment, which is reflected by the similarity of their manifestos with those of older parties in regard to populism. 50# 45# 40# 35# Degree#of#populism# 30# 20# 10# 0# 13# 13# 13# 13# 27# 18# 16# 14# 13# 11# 4#4# PDA# ( )# 32# 10# 2# 2# POCH# ( )# 6# 3# 3# 4# 4# 4# 4# 3# 3# 3# 3# 3# 3# 3# 2# 3# 1# SD# ( )# GP# ( )# 15# 10# 8# 5# FP# ( )# 5# 1# 0# 0# 0# 0# 0# EDU# ( )# 1.#elecGon# 2.# 3.# 4.# 5.# 6.# 7.# 8.# 9.# 10.# 11.# 12.# 13.# 14.# Figure 5. The degree of populism of 6 parties, which entered the party system during the research period. The GLP and the BDP - which took part in the elections only since 2007 and 2011 respectively - had only populist values of 0 and are not presented in the figure. The first bar for each party represents the value at the time, when the party for the first time gained at least 1% in the national elections. The dotted line represents the average value between of those 6 parties, which were already part of the political system before the beginning of the research period (2.67). Figure 5 shows, that the POCH and to a somewhat lesser degree the FP downed down their popu- list discourse during the 20 years of their existence on the Swiss political landscape. This is much in 12

13 line with the third hypothesis, which states that new parties having a strong populist discourse inte- grate into the power structure of the state and as a consequence become less populist. A similar de- velopment can be stated for the LDU, which was founded only shortly before the beginning of the research period (see figure 6). While the PDA increased its already high degree of populism until the 1970s and became only afterwards less populist, SD and GP are also against the expectations of the third hypothesis - not becoming less populist over time. The data are thus too unclear and diverse to clearly answer the third hypothesis 40" 30" 30" Degree"of"populism" 20" 19" 17" 17" 10" 0" 10" 10" 8" 8" 8" 7" 6" 6" 6" 6" 5" 5" 5" 5" 5" 5" 5" 4" 3" 3" 4" 3" 2" 2" 3" 2" 3" 2" 2" 3" 3" 2" 1" 2" 2" 2" 2" 1" 2" 2" 2" 1" 1" 2" 2" 1" 2" 1" 1" 1" 1" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" 0" SP" " FDP" " CVP" " SVP" " EVP" " LDU" " 1951" 1955" 1947" 1959" 1963" 1967" 1971" 1975" 1979" 1983" 1987" 1991" 1995" 1999" 2003" 2007" 2011" Figure 6. The degree of populism of those 6 parties, which were already part of the political system before Conclusion The rise of the new populist Right in Western Europe is in the scientific literature often affiliated to the process of denationalization of politics. The national political scope of action and therefore also popu- lar sovereignty has been reduced in the process of globalization, which serves as a breeding ground for populism. However, there is no empirical analysis about the development of the degree of popu- lism in the political discourse, which covers more than the last two decades. Considering the time pe- 13

14 riod from , this study thus analysis if, in the case of Switzerland, parties have been more populist in the last years than before or if the emergence of the new radical right parties with their strong populist discourse is only the newest piece of a constant renewal of the party systems. The results indicate that in the case of Switzerland parties were on average not more populist in the last two decades than before. Since the beginning of the after war period there were always mainly new parties with a strong populist discourse in their election manifestos. The Swiss People s Party (SVP), as typical representative of the new populist right in Western Europe, is therefore no exception in this respect. Although, the SVP is electorally clearly more successful than other strongly populist parties of the past and we can therefore state an increase in the degree of populism in the last years if we weight the parties election manifestos according to their electoral support. But since this increase disappears if we drop the SVP from the analysis we can not speak of a general populist Zeitgeist. Un- der the surface of a relative constant average degree populism in the election manifestos during the research period, it was however possible to clearly show that while the populist discourse was at the beginning of the research period dominated by left parties, radical right parties constantly gained ground in this respect since the 1970 and are today clearly more populist than their left counterparts. While I believe that this study is able to highlight the broad development in the degree of popu- lism of the whole Swiss party system over time, it has some limitations concerning the analysis of indi- vidual parties populism. The data basis of one election manifesto per party every 4 years is probably too rough- grained for the analysis of the development of a specific party s populist discourse over time (hypothesis 3). This seems to me a more problematic aspect concerning the use of election mani- festos as data source, than the general question on how well they represent a party s ideology in gen- eral or its populist character more specifically (cf. Rooduijn et al. 2012). Above that, it is of course not clear how representative these findings of the Swiss case are for other Western European countries. The fact that the emergence of a strongly populist and highly successful radical right party in the last years is in line with developments in other long established democracies seems to indicate however, that Switzerland is not a extraordinaire case in respect to the development of populism over the last decades. 14

15 Notes 1 Rooduijn et al. (2012) also use a longitudinal design to study the development of populism over time. However, 2 In the populist literature there are some contestations concerning the understanding of populism as an ideology. First, populism is sometimes rather seen as a communication strategy, which is only instrumentally adopted for electoral reasons (e.g. Jagers and Walgrave 2007). But this distinction between ideology and strategy seems to be rather unhelpful in an analysis of populist discourse. It is nearly never possible to definitively decide whether parties use of populism is motivated by electoral opportunism or by an underlying ideology (March 2012: 5). Moreover, populism is sometimes associated with specific communication styles like simplistic or emotional political language etc. (e.g. Taggart 2000, Jagers and Walgrave 2007). In using a colloquial style, political actors show that they are close to the people and speak their language (March 2012: 4f). The analysis of this paper is however restricted to the content of the political messages and does not measure different styles in which this contents are presented. 3 Political scientists often disaggregated the former aspect of the antagonistic relationship between the people and the elites into two separate elements. These are usually labelled as people-centrism (the mentioning of the people as a unified actor) and anti-elitism (e.g. March 2012, Kriesi 2013, Rooduijn 2013, Rooduijn et al 2012). Accordingly, Rooduijn et al. (2012) measured people-centrism and anti-elitism separately in their content analysis. However they acknowledged that the presence of only one of these elements is not enough for a statement to be coded as populist. It is the combination of people-centrism and anti-elitism that defines populism (ib.: 7). They thus defined a statement to be populist in case both elements are present in the same paragraph, implicitly assuming that if the speaker criticizes the elites and make a reference to the people, also speaks about their (antagonistic) relationship. I would argue that it is more valid to directly searching for statements about the antagonistic relationship between people and elites. 4 As is well known the Swiss political system is not only a special case regarding the number of its parties but also because of its direct democratic and consensual institutions. It is however still unclear what effects they have on the degree of populism (cf. Papadopoulos 2002: 53). On the one hand, the possibility of initiatives and referendums and the absence of a constitutional court bring politics closer to the people, which is why Kriesi and Gentile (1998: 129ff) argued that the existence of direct democratic channels had somehow pre-empted populism in Switzerland in the past. Above that, with its consensus democratic system, Switzerland is able to include all large parties and therefore all relevant societal groups in the decision making process (Trechsel and Sciarini 1998) which should further reduce the possibility that substantial parts of the population feel excluded from the political process. On the other hand, several authors assume that in countries with broad government coalitions, radical political parties are more likely to criticize the whole political elite for not serving the people (e.g. Skenderovic 2007: 57). And because Switzerland is only partly direct democratic there are always tensions between the direct democratic and the representational pillar of Swiss democracy. On the whole, the political system in Switzerland has a high representativeness and responsiveness, but is at the same time less accountable (Kriesi 2013:210). If Switzerland constitutes a case where the political discourse generally contains rather an above- or below-average 15

16 case compared to other Western European countries is thus difficult to answer theoretically and there is no empirical study yet about this question. 5 Three smaller parties the republican Movement; the Democratic Party and the Liberal Party would also fulfil the stated criteria to be selected in this study. Since for this parties, which today all are dissolved, no election manifestos are available, they cannot be included in the analysis. By contrast, the Democratic Party of Switzerland (BDP) was considered in this study even they only took part in one elction yet. This is because of their relative high vote share of 5.4% in the last elections of the national parliament. 6 The only exceptions are the absence from government by the Social Democrats (SP) between 1953 and 1959 and of the Swiss people s party (SVP) in Moreover, the new Democratic Party of Switzerland (BDP) is part of the national government since To a higher degree than in other Western European democracies, the political system in Switzerland has with the exception of the more centralized social democratic and communist parties (Gruner 1977: 29) - always been regionally shaped. And despite the recent changes towards a nationalization of the political landscape (Caramani 2004), the cantonal organizations are still important actors (Mazzoleni 2005: 212). 8 Strictly following the weighting of Rooduijn et al. (2012: 16), I first computed the Z-scores for every manifesto based on the mean length of the manifestos (measured by the number of paragraphs). If the Z-score was below -2 or above 2, I weighted the populist paragraphs with the factors 0.5 and 2 respectively. If they were between -1 and -2, I gave a weight of 0.75, if it was between 1 and 2, the weight is In 18 out of 150 cases there was no election manifesto available for one of the considered parties in a specific election period. Because of the extensive research for the manifestos in party as well as cantonal and national archives, this is mostly the case when there actually never existed a manifesto because the respective party decided not to release a new one. For the analysis of this study, the populist scores of this data points are defined as being equal to those of the preceding election manifesto of the respective party. 16

17 References Albertazzi, D., & McDonnell, D. (Eds.). (2008). Twenty- first century populism: the spectre of western European democracy. Houndmills, Basingstoke, Hampshire [England] ; New York: Palgrave Macmillan. Canovan, M. (1999). Trust the People! Populism and the Two Faces of Democracy. Political Studies, 47(1), Canovan, M. (2002). Taking Politics to the People: Populism as the Ideology of Democracy. In Y. Mény & Y. Surel, Democracies and the Populist Challenge (pp ). New York: Palgrave. Caramani, D. (2004). The nationalization of politics: the formation of national electorates and party systems in Western Europe. Cambridge, UK ; New York: Cambridge University Press. Dahl, R. A. (1994). A democratic dilemma: system effectiveness versus citizen participation. Political Science Quarterly, Decker, F. (2006). Die populistische Herausforderung. Theoretische und ländervergleichende Perspektiven. In F. Decker (Ed.), Populismus (pp. 9 32). VS Verlag für Sozialwissenschaften. Gruner, E. (1977). Die Parteien in der Schweiz, Bern. Jagers, J., & Walgrave, S. (2007). Populism as political communication style: An empirical study of political par- ties discourse in Belgium. European Journal of Political Research, 46(3), Kaltwasser, C. R. (2011). The ambivalence of populism: threat and corrective for democracy. Democratization, 19(2), Kriesi, H. (2013). Conceptualizing the populist challenge. Presented at the John Hopkins University, Bologna. Kriesi, H. (2014). The Populist Challenge. West European Politics, 37(2), Lijphart, A. (1999). Patterns of Democracy: Government Forms and Performance in Thirty- six Countries. Yale University Press. Mair, P. (2002). Populist Democracy vs Party Democracy. In Y. Mény & Y. Surel, Democracies and the Populist Challenge (pp ). New York: Palgrave. March, L. (2007). From Vanguard of the Proletariat to Vox Populi: Left- populism as a Shadow of Contemporary Socialism. SAIS Review, 27(1), March, L. (2012). Towards an understanding of contemporary left- wing populism. Presented at the Political Stud- ies Association (PSA), Annual International Conference, Belfast. March, L., & Mudde, C. (2005). What s Left of the Radical Left? The European Radical Left After 1989: Decline 17

18 and Mutation. Comp Eur Polit, 3(1), Mazzoleni, O. (2005). Multi- Level Populism and Centre- Periphery Cleavage in Switzerland The Case of the Lega dei Ticinesi. In D. Caramani & Y. Mény (Eds.), Challenges to Consensual Politics Democracy, Identity and Populist Protest in the Alpine Region (pp ). Brussels: P.I.E Peter Lang. Mény, Y., & Surel, Y. (Eds.). (2002). Democracies and the populist challenge. New York: Palgrave. Mudde, C. (2004). The Populist Zeitgeist. Government and Opposition, 39(4), Mudde, C. (2007). Populist radical right parties in Europe. Cambridge, UK ; New York: Cambridge University Press. Pauwels, T. (2014). Populism in Western Europe: comparing Belgium, Germany and the Netherlands. Milton Park, Abingdon, Oxon ; New York, NY: Routledge. Rooduijn, M. (2013). The Mesmerising Message: The Diffusion of Populism in Public Debates in Western Europe- an Media. Political Studies. Rooduijn, M., de Lange, S. L., & van der Brug, W. (2012). A populist Zeitgeist? Programmatic contagion by popu- list parties in Western Europe. Party Politics. Skenderovic, D. (2007). Immigration and the radical right in Switzerland: ideology, discourse and opportunities. Patterns of Prejudice, 41(2), Taggart, P. (2000). Populism. Buckingham [England] ; Philadelphia: Open University Press. Taguieff, P.- A. (1995). Political Science Confronts Populism: From a Conceptual Mirage to a Real Problem. Telos, 1995(103), Trechsel, A. H., & Sciarini, P. (1998). Direct democracy in Switzerland: do elites matter? European Journal of Political Research, 33(1),