1
Introduction
In their seminal work, Lipset and Rokkan (1967) argued that the party systems of contemporary European democracies were frozen. According to the authors, all political systems were structured around deep and enduring cleavages such as the opposition between the centre and periphery, church versus state and labour versus capital. And since most party families were positioned along these stable cleavages, not much change had to be expected in the party systems of Western Europe. Yet only a few years after their path breaking work, the first signs of a defreezing political system became visible. In the 1960s and 1970s environmental parties emerged in many democracies and as a reaction, numerous parties of the radical right gained momentum since the 1970s (Ignazi, 1992). Other parties, which are more difficult to classify, also started challenging the established parties. Examples of these âhybridsâ are the Fremskridtspartiet (Progress Party, FPd) in Denmark, the Italian Forza Italia (Go Italy, FI) or the Socialistische Partij (Socialist Party, SP) in the Netherlands.
What is remarkable about these new parties is their diversity, in terms of both ideology and their ability to obtain and maintain electoral success. A substantial number of these challengers has been grouped together, however, under the banner of populism (MĂŠny and Surel, 2002). It is argued that some of them share a populist core as they all claim to defend the âpure peopleâ against the âcorrupt eliteâ while arguing that politics should be an expression of the general will of the people (Mudde, 2004). In countries around the globe ranging from Venezuela (Hugo ChĂĄvez) to Italy (Lega Nord) â populist challengers have had a profound impact on the political system. Even in countries with a reputation for tolerance like the Netherlands, Sweden and Finland, populist politicians have achieved electoral success by denouncing immigration, the European Union and the ruling political elites in the last few years. Since populism is often conceived of as being the new spectre of democracy, in Western Europe as well as elsewhere, the research on the topic has proliferated lately (Albertazzi and McDonnell, 2007).
Existing research
One stream of literature has focused on the conceptualization of populism (Ionescu and Gellner, 1969; Mudde, 2004; Taggart, 2000). For more than half a century, scholars have been debating on what populism exactly is and how it should be defined. Is populism a mobilization strategy, a political style or an ideology? Even though this appears to be a frustrating endeavour, it is nonetheless essential for the accumulation of knowledge to start from a common conceptual ground. In recent years it seems that some kind of consensus has been found among researchers, in that the political exploitation of an alleged division between âthe peopleâ and âthe eliteâ is what populists thrive on and enables identifying populist actors (Mudde, 2007; Rooduijn, 2013). By contrast, the measurement of populism has not received much attention. The term is often attached to certain parties without systematically exploring their ideologies. Nor are the criteria to label a party as populist always specified. Here, too, research has been expanding recently, drawing on increasingly sophisticated methods such as content analysis to measure populism (e.g. Hawkins, 2009).
Another topic that received considerable attention is the relationship between populism and democracy (Abts and Rummens, 2007; Rovira Kaltwasser, 2011). Both theoretically and empirically, scholars have tried to answer the intriguing question whether populism is a pathological phenomenon leading to a degeneration of democracy or whether, on the contrary, populism might be a more authentic form of representation. Even though this debate is far from settled it seems that recent research increasingly acknowledges that, in some circumstances, populism might be a corrective to democracy (Mudde and Rovira Kaltwasser, 2012).
Turning to the more empirical work on populism, scholars have been focusing on how populist parties deal with government participation (Albertazzi, 2009; Delwit, 2007; Heinisch, 2003). These studies have yielded interesting results and the idea that populist parties are only successful in opposition but have difficulties once in government has gradually been abandoned (Albertazzi and McDonnell, 2010). More generally, it seems that populist parties are often more resilient than expected and have become increasingly anchored in contemporary party systems. Another promising lead of investigation in populism studies deals with the measurement of populist attitudes and to what extent these are supported by citizens (Hawkins, Riding and Mudde, 2012; Stanley, 2011).
Despite the many different avenues in populism research, most studies to date still focus on explaining the rise of populist parties (Albertazzi and McDonnell, 2007; Ivarsflaten, 2008). The problem with populism is, however, that it rarely exists in isolation. Populism is almost always attached to another ideology which makes the understanding of its nature and causes particularly complex. Consequently, we have learned a lot about the rise of populist radical right parties (Arzheimer, 2009; Carter, 2005; Golder, 2003; Jesuit, Paradowski and Mahler, 2009) and to a lesser extent about left-wing populist parties (March and Rommerskirchen, 2012; Weyland, 2003) yet we know remarkably little about general causes of populist party success (for exceptions, see Hawkins, 2010; Hino, 2012). Whereas numerous studies found that anti-immigrant attitudes at the individual level increase the likelihood of populist radical right voting (e.g. Ivarsflaten, 2008), it is unclear whether this is related to the host ideology (i.e. ethnic nationalism) or the populist appeal of these parties. The few studies that have focused on a wide variety of populist parties have pointed to broad factors such as the convergence of mainstream parties, the media and globalization (Albertazzi and McDonnell, 2007), yet a systematic and more detailed empirical verification of these theories is still lacking.
Challenges in explaining the success of populist parties
At least three problems hamper our understanding of populist party success. The first has been touched upon in the previous paragraph. The abundance of studies on the populist radical right do not allow us to determine which factors are related to the populist character of these parties and which to other features. Possibly there exist some explanatory variables that are similar for the rise of all populist parties. An investigation into which factors (if any) are common for populist parties might be illuminating in this respect. Second, Mudde (2007: 29) suggests that we should distinguish between the populist radical right, social populist and neoliberal populist parties.1 The first group contains âprophetsâ â addressing the new issue of immigration â while the latter two are âpurifiersâ claiming to defend the ârealâ socialist or (neo)liberal values which have allegedly been betrayed by the established parties (Lucardie, 2000). Again, while there is much research on the populist radical right, it is largely an open question what explains the rise of neoliberal and social populist parties. Although the latter are less numerous and successful (in Europe), some of them have become serious competitors for established parties. A third shortcoming in existing research is the focus on the breakthrough of these parties while less is known about why parties persist or not. As most research is cross-national in nature or focused on the individual voter, it does not provide answers to the question why there is often a considerable variation in populist party success over time. This is particularly disturbing since many populist parties have been very unstable in their electoral performances. The âflashâ performances of parties such as Lijst Pim Fortuyn (List Pim Fortuyn, LPF) in the Netherlands or the Swedish Ny Demokrati (New Democracy, ND) demonstrate this.
This study pushes forward our understanding of populist parties in three ways. First, I will propose a typology of populist parties and classify parties in three West European countries according to this typology. This typology has the advantage of not having to resort to the too generic term âpopulist partiesâ. Second, it will be investigated which socio-demographic characteristics and attitudes contribute to populist voting in general at the individual level. It is explored who votes for populist parties and for what reasons by drawing on a comparative research design. The general aim of this study is to explore and compare the voters of a wide variety of populist parties in search for a common denominator. Third, this study will shed light on why some populist parties are persistently successful while others have disintegrated fast. This will be done by focusing on factors such as organizational characteristics and leadership.
Research question
The basic research question guiding this book is the following: What causes people to vote for a populist party? It is expected that, depending on which ideology populism is combined with, these parties will attract different kinds of voters. By studying the electorates of a range of different populist parties, I attempt to disentangle what is exactly the populist element, rather than elements related to the host ideology, that drives voters towards these parties. The underlying logic is that explanatory factors such as political distrust or opposition to immigration have the same impact on different subtypes of populist parties but not necessarily on all populist parties. Those variables that do have the same impact on all cases can be related to the populist character of the party and not to its adjacent ideological characteristics.
Before the variety of populism can be explored, several related questions have to be dealt with. The first is concerned with the definition of populism. Since populism is a contested concept, this issue deserves considerable attention. What do we mean with the term âpopulismâ? The most recurrent meanings of the concept will be reviewed in an attempt to arrive at a systemized concept. But even in a more substantive form â populism as a thin centred ideology â it is commonly acknowledged that populism needs another ideology to attach itself to (Deegan-Krause and Haughton, 2009). This means that a considerable part of the identity of populist parties is derived from other ideological components. Consequently, it seems necessary to build a typology of populist parties. The typology of Mudde (2007), who distinguishes between neoliberal populism, social populism and the populist radical right, provides a useful starting point but will be refined and elaborated.
Another important question is that of classification: Which parties can be labelled as populist? Or more precisely: Which parties are populist radical right, neoliberal populist and social populist? Since the measurement of populism has received scant attention, it remains a challenge to arrive at a systematic and valid classification. In this study, I will examine which parties can be labelled as popu-list in the party systems of Germany, Belgium and the Netherlands. These countries have been selected because existing research suggests that they host a variety of populist parties. Another advantage of taking these countries into consideration is that they share a similar institutional setting. This allows for a most similar system design, making the impact of other causal factors such as the electoral system rather limited (Przeworsky and Teune, 1970).
Probably the rarest subtype of populist parties in Europe is the left-wing variant. Within Western Europe, March and Mudde (2005) only identified the Scottish Socialist Party (SSP), the Dutch SP and some small radical left movements in France (e.g. Lutte Ouvrière) as social populist. âIn Eastern Europe, the East German PDS would be the ideal typeâ (March and Mudde, 2005: 36). Since the latter transformed and renamed itself to Die Linke (The Left, DL) and is currently competing for votes among the whole of Germany, this seems also a case worth considering. Neoliberal populist parties are not very common either. Among the most known are probably the Dutch LPF, the Danish FP, FI in Italy and the Belgian Lijst Dedecker (LDD) (Mudde, 2007; Pauwels, 2010). Populist radical right parties are far more numerous and successful than the other two subtypes. The most known in Western Europe are the Belgian Vlaams Belang (Flemish Interest, VB)2, the Austrian Freiheitliche Partei Ăsterreichs (Austrian Freedom Party, FPĂ), the Dansk Folkeparti (Danish Peopleâs Party, DFP) in Denmark, Die Republikaner (REP) in Germany and the Front National (National Front, FN) in France.
Finally, after a careful classification and measurement of populism, I will turn to the main research question by exploring who votes for populist parties and why. Drawing mainly on the sociological model of voting it is hypothesized that certain social groups are more likely to vote for populist parties than others. While some hypotheses are expected to be relevant for all parties under study, others will be differentiated according to the type of populist party. In a next step, the socio-demographic characteristics and attitudes of the voters for different kinds of parties will be examined. For each case it will also be explored to what extent the party has been able to persist over time. This sheds light on why certain parties are here to stay while others have faded away quickly.
Data and methods
Two main data sources will be used to answer the research questions previously outlined. First, in an attempt to classify political parties in accordance with the proposed typology, I will use party programmes as main data sources. These documents are commonly used to explore the policy positions and ideologies of political parties (Budge, Klingemann, Volkens, Bara, and Tanenbaum, 2001). Since it is argued, moreover, that populism can be best conceptualized as a thin centred ideology it should be possible to identify populism in party manifestos.
Two methods will be combined to identify populist parties in different party systems. The first is a qualitative analysis on the basis of party ideology (i.e. party programmes and secondary literature) (see Mair and Mudde, 1998; Mudde, 2000). Investigating for each party whether it matches a minimal definition of populism allows for a Sartorian classification of parties. However, this qualitative approach leaves a lot of room for interpretation to the researcher. Given the debate on classification (exacerbated by the negative connotation of populism) and the existence of borderline cases, the obtained results will be complemented wit...