The Triumph of (Underlying)Ideology Over Populism in Western Europe

The prototypical form of populism in Europe has been that of the radical right, which combines populism with nationalism, xenophobia and certain doses of authoritarianism. European left-wing populism, for its part, had remained a marginal phenomenon until the outbreak of the economic crisis in 2008. Since then, populism has ceased to be a phenomenon almost exclusively for the radical right and has spread along the ideological spectrum or has appeared with ambiguous ideological positions. The recent electoral advances of populism in Europe have led to the formation of coalition governments between populist parties of different ideological signs (first in Greece, then in Italy). Likewise, the programmatic evolution followed by some populist parties (e.g., the populist radical right's shift to the economic center, or even center-left) or some similarities between these parties beyond their populist rhetoric (e.g., Euroscepticism), indicates that European populist parties may have more in common than might be expected. This leads us to the following question: Are we witnessing the triumph of populism over ideology? That is, do left and right populist parties tend to converge on other issues that are beyond their populist rhetoric? Or do left-right differences remain hegemonic? This article aims to contribute to a better understanding of the nature of populism in Europe. In particular, this article aims to determine whether underlying ideology triumphs over populism in these types of political organizations or not. In order to do that, this study will analyze the ideological positioning and cohesiveness of populist parties in Western Europe at both party and electorate levels. Therefore, the ultimate goal of this research is to shed light on a phenomenon that is advancing electorally in Europe and that could determine future coalitions and government alliances.


INTRODUCTION
Radical right-wing populism has been the prototypical form of populism in Europe (the Austrian Freedom Party, the Dutch Party for Freedom, the Swiss People's Party, etc.). However, the Great Recession of 2008 led to the rise of radical left-wing populist parties, especially in those countries hardest hit by the crisis and austerity measures (e.g., Podemos in Spain, Syriza in Greece) as well as "post-ideological" or centrist populist parties (e.g., M5S in Italy, ANO2011 in Czechia) that have challenged the traditional left and right divide. The recent electoral advances of populism in Europe have even led to the formation of coalition governments between populist parties of different ideological signs (first in Greece, then in Italy). Likewise, the programmatic evolution followed by some populist parties (e.g., the populist radical right's shift to the economic center, or even centerleft) or some similarities between these parties beyond their populist rhetoric (e.g., Euroscepticism), suggests that European populist parties may have more in common than might be expected. This leads us to the following question: Are we witnessing the triumph of populism over ideology? That is, do left and right populist parties tend to converge on other issues that are beyond their populist rhetoric? Or do left-right differences remain hegemonic? The answer to these questions could have different implications for future government alliances in Europe as well as for the evolution of European democracies. The "populism-trumps-ideology" thesis (March, 2017) would imply that we are facing a gap in the political space that pits "populists" against "anti-populists." In this case, future government alliances could be forged on the basis of this criterion (as happened in Italy with the coalition of Lega and M5S, or in Greece, with Syriza and ANEL), with all that it implies for the European integration project and for democratic stability. On the contrary, the "ideology prevails over populism" thesis would mean that left-right differences prevail over the (anti)populist agenda of European political parties. In this case, government or parliamentary alliances would generally follow the patterns of ideological bloc politics, pitting right against left parties (as in Spain since 2018 with the confrontation of the two ideological blocs: PSOE-Podemos vs. PP-Cs-Vox).
Therefore, this article aims to contribute to a better understanding of the nature of populism in Europe. In particular, this research aims to determine whether underlying ideology triumphs over populism in these types of political organizations or not. In order to do that, this study will analyze the ideological positioning and cohesiveness of populist parties in Western Europe at both party and electorate levels. The ultimate goal of this research is to shed light on a phenomenon that is advancing electorally in Europe and that could determine future coalitions and government alliances.

THEORY
The chameleon-like nature of populism has led to an intense theoretical discussion of this phenomenon. Indeed, populism has been approached as ideology, strategy and discourse (Taggart, 2002). In this article, the nature of populism is discussed in relation to the political ideas that accompany it. As March (2017) notes, it is possible to identify two general approaches to populism in this regard: the first approach highlights the ideological distance between the different types of populist parties. According to this approach, "ideology trumps populism (i.e., right-left differences remain hegemonic)" (p. 285). The second, on the contrary, emphasizes the similarities between populist parties. It considers that populism trumps (underlying) ideology, that is, that "right and left are essentially similar qua populist parties" (p. 284).
In the first approach we could include those authors who, from a normative-Laclauian perspective, argue that the progressive and emancipatory agenda of left-wing populism make it not only fundamentally different from right-wing populism, but a clear adversary of it (Mouffe, 2016). In this regard, Podemos, a party clearly inspired by the ideas of Mouffe and Laclau (Franzé, 2017), warned in the electoral program for the 2019 European elections of "the emergence of a reactionary axis, led by Salvini, Orban and Le Pen, now joined by the Spanish right-wing with the intention of liquidating European values and identity at the pace of an authoritarian, xenophobic, homophobic and misogynist discourse 1 " (p. 5).
In addition to this normative perspective, the ideational approach to populism implies that host ideology should matter more than populism, a thin ideology that presents a restricted morphology based on a small number of central concepts whose meaning is context-dependent (Mudde and Rovira Kaltwasser, 2013). The central concepts of populism are the conception of the people as a virtuous entity, the demonization of the elites and the supremacy of the popular will. These would be the common elements of populist parties. However, the categories of "people" and "elites" should be considered as empty vessels whose substantive content depends on the context and the underlying ideology of the populist actors (ídem).
In the case of the European populist right, Mudde (2007) argues that the term "populist radical right" is more appropriate than "radical right populism" since "nativism, not populism, is the ultimate core feature of the ideology of this party family" (p. 26). Nativism is defined by Mudde (2007, p. 19) as "an ideology, which holds that states should be inhabited exclusively by members of the native group ('the nation') and that non-native elements (persons and ideas) are fundamentally threatening to the homogenous nation-state." The definition of nativism as the ideological core of the populist radical right implies at least two things: first, that the anti-immigration and nationalist agenda is more important for the populist radical right than the populist agenda, and second, that the categories of the people and the elites are determined by its nativist orientation: "the people" refers to those who are part of the nation, excluding those who are outside of it. Likewise, the anti-elitist discourse is directed against those powerful groups-in the sphere of politics, the media, culture, etc.-that threaten the integrity or the interests of the nation. As Mudde (2007) argues, the basis for defining "nativeness" can be diverse (e.g., ethnic, racial, religious, etc.) but always have a cultural component. For this reason, these parties tend to compete fundamentally in the cultural dimension to the detriment of the economic dimension (Fernández García, 2019) 2 .
As regard the European populist left or "social populists, " they retain a democratic socialist ideological core although "they generally have far less concern with doctrinal purity and class-consciousness than the traditional left" (March, 2007, p. 66). The defense of collective economic and social rights and economic redistribution constitutes their main agenda. This socialist-democratic ideological core explains the socioeconomic orientation of the people they claim to represent (i.e., "those at the bottom") as well as the elites they attack (powerful economic groups, banks, "the Troika, " etc.). In this regard, while the populist radical right considers that the elite has betrayed the 1 Translated by the author from Spanish. 2 More specifically, the main issues on which European right-wing populist parties compete are immigration and nationalism (nativism, if we combine it) while the main issue for left-wing populist parties is redistribution (Fernández García, 2019, p. 165). people by introducing non-native elements into their societies (both people -immigrants and refugees-and ideas or values -Islam and multiculturalism-), the populist left considers that the people have been betrayed by an elite which serves neo-liberal capitalistic interests.
Therefore, although this approach identifies important similarities between populist parties on the right and the leftsuch as anti-elitist rhetoric, emphasis on popular sovereignty and attacks on the liberal pillar of democracy (Mudde and Rovira Kaltwasser, 2013)-it emphasizes the ideological differences between them. These differences are explained by the underlying ideology of populist actors. The first and most important difference is the exclusionary character of the populist right compared to the inclusionary nature of the populist left. This exclusionary or inclusive character refers to their conception of the people, in the sense that the populist right tends to exclude those groups that are outside the nation (according to them, immigrants, refugees, ethnic or religious minorities). In a vertical dimension, however, both types of populisms are exclusionary since they exclude the elites from their conception of the people. In this regard, the populist left is dyadic (attacking elites), while the populist right is triadic (attacking elites and out-groups). The second difference is that while the populist right is predominately focused on ethnic, cultural or national identity, the populist left tends to focus on economic issues (March, 2017). This approach to populism is supported by many empirical analyses at both the party and electorate levels. At the party level, March's study (2017) of the British populist parties concludes that "host ideology is more important than populism per se in explaining the essence of left and right-wing populisms" (p. 300). This case study shows that although left and right populisms in the United Kingdom share specific populist orientations (antielitism, people-centrism and advocating popular sovereignty), they maintain substantial differences that are explained by their underlying ideologies: the BNP's populism is mostly ethnocultural (indigenous people vs. left-liberal elite), the UKIP's is economic (neo-liberal) and cultural (taxpayers/citizens vs. bureaucrats/cultural elites), and both the SSP and Respect's is economic (working people vs. capitalists). The comparative analysis of Fernández-García and Luengo (2018) reaches a similar conclusion: populist parties on the right and left agree on identifying a conflict between the powerful elite and ordinary people but differ on the reasons that lead to sustaining this antagonistic view of society, as well as on the targets of their populist discourse. In addition to the political establishment, the populist right tends to attack foreign elites and liberal intellectuals for promoting European integration and multiculturalism, phenomena that threaten the integrity of the nation according to this type of parties. The populist left, for its part, concentrates its attacks on economic powers and ruling elites for threatening the (economic) sovereignty and social rights of the peoples.
The argument that underlying ideology trumps populism also finds empirical support in Otjes and Louwerse's (2015) research on the behavior of Dutch populist parties in political institutions. These authors conclude that left-right position is more important than populism when it comes to vote in Parliament: the populist Socialist Party votes in a similar way to the other parties of the left while the Party for Freedom does as the other parties on the right. In the case of the populist right, for example, its anti-immigration agenda is more important than its populism, supporting Mudde's argument that nativism, not populism, is the ideological core of these parties.
If we observe the transnational party coordination in the European institutions, we can reach a similar conclusion. In the current legislature of the European Parliament, the main radical right-wing populist parties have joined the Identity and Democracy group, which succeeded the Europe of Nations and Freedom Group. It has 73 members from nine countries (it is the fifth largest group in the Parliament), with the majority coming from Italy's Lega party, the National Rally in France and Germany's AfD. The rest of the members of this group are the Freedom Party of Austria, the Flemish Interest, the Finns Party, the Danish People's Party, the Conservative People's Party of Estonia and the Czech Freedom and Direct Democracy. Other radical right-wing populist parties such as the Swedish Democrats, the Spanish Vox or the Polish Law and Justice, are part of the European Conservatives and Reformists Group, another group oriented to the right. With regard to the populist left, they are organized in the Confederal Group of the European United Left -Nordic Green Left (e.g., France Insoumise, the Belgian PVDA-PTB, the Greek Syriza, the Spanish Podemos, the Irish Sinn Féin, the Portuguese Left Bloc, etc.). The fact that populist parties are organized at the European level on the basis of their left-right ideological orientation rather than their populist orientation reinforces the argument that underlying ideology prevails over populism in Europe. Especially if we take into account that Euroscepticism is precisely one of the elements that brings these populist parties closer together, and could therefore be an element that unites them in European institutions. In this regard, the 5 Star Movement, "a political party that "does not articulate any 'full' ideology, but rather expresses a purely ('thin') populist ideology" (Manucci and Amsler, 2018, p. 127), is in the non-attached group in the EP, which reinforces this argument.
At the electorate level, we also find support for this thesis. The case study of Dutch populist parties of Akkerman et al. (2017) shows that while supporters of both the PVV and the SP do not significantly differ in their populist stances, they do differ in their attitudes toward immigration and the economy. In this regard, the populist right seems to attach to a cultural ideology of exclusion, while the populist left attaches to an economic ideology of inclusion. Similar findings are found in the cross-national analysis of Van Hauwaert and Van Kessel (2018): populist attitudes are significant predictors of the support for both left and right-wing populist parties, but they substantially differ in the cultural and economic dimension. Those who hold left-wing economic positions and are culturally liberal are more likely to support left-wing populist parties while individuals with authoritarian preferences and anti-immigrant attitudes are more likely to support right-wing populist parties. Finally, the comparative study of Rooduijn (2018) concludes that there is no such thing as a populist voter base: not all populist voters are distrustful and Eurosceptic "losers of globalization" who demand more direct democracy. In line with the previous studies, this analysis shows important differences between voters of both types of populist parties with regard to immigration: voters who hold negative attitudes toward immigrants are more likely to vote populist only when it comes to radical right-wing populist parties.
Therefore, there is a broad empirical basis that supports the argument that populist parties in Europe share a populist conception of democracy and politics, but differ substantially in the ideological positions they hold. In general, the populist right is exclusionary and competes mainly in the cultural dimension, emphasizing issues of national, cultural, ethnic or religious identity. In contrast, the populist left tends to be inclusive, liberal in the cultural dimension, and focused on socio-economic issues, emphasizing values such as social justice and equality. However, we also find empirical studies that question this supposedly clear differentiation between left-and right-wing populisms, leading us to question whether ideology really trumps populism.
First, the study of Rama and Santana (2019) challenges the results of the previously cited election studies by showing that nativist attitudes do not affect the left vs. right populist competition as one might expect. According to the authors, this finding is consistent with the development of some left-wing populist parties (e.g., in Italy, Germany or the Netherlands) by which they have adopted common right-wing populist discourses against immigration.
Second, the study of Halikiopoulou et al. (2012) identified an increasing party polarization on issues of cultural identity and European integration independent of the left-right dimension. The authors concluded that radical right and left populist parties "side together on the axis measuring opposition to/support of European integration as well as on the dimension measuring levels of nationalism" (p. 531). In a similar vein, the study of Rama and Santana (2019) also shows that detachment from Europe increases the likelihood of voting a populist party, right and left being indistinguishable. However, as Halikiopoulou et al. (2012) pointed out, right-wing populist parties express their nationalist and Eurosceptic stances from a predominantly ethnic viewpoint while left-wing populist parties adopt a predominantly civic perspective. In this regard, the study of Plaza-Colodro et al. (2018) shows that populist parties indeed share a Eurosceptic profile but a further examination indicates that their positions toward the EU are mediated by their thick ideology: rightwing populist parties' Euroscepticism is directed against the foundational pillars of the EU (i.e., integration and borders) while left-wing populist parties focus their opposition to the economic dimension of the organization.
Related to nationalism and opposition to the EU, it has been pointed out that economic protectionism is another area where populist parties tend to converge. Opposition to trade opening has been associated with left-wing populist parties as a "part of a more encompassing economically egalitarian agenda that includes other policies aimed at mitigating the inequalities resulting from free trade, such as welfare arrangements and a tax system that redistributes income from the rich to the poor" (Van der Waal and De Koster, 2018, p. 562). One of the most recent episodes of trade protectionism displayed by European populist parties has been the strong opposition to the Transatlantic Trade and Investment Partnership. For example, the 2015 Podemos' manifesto for the general elections stated that these kinds of agreements threaten "our sovereignty, our democracy, our economy and our Welfare State 3 " (p. 221). The leader of the National Rally, Marine Le Pen, also said on Twitter: "The #TTIP/#TAFTA denies the differences between peoples and the diversity of their identities 4 ." Indeed, we also find some rightwing populist parties displaying protectionist positions, such as the already mentioned National Rally in France and the Party for Freedom in the Netherlands. In this regard, the study of Van der Waal and De Koster (2018) concluded that protectionism serves as a determinant of voting for a populist party whether on the right or the left. However, as the same authors point out, the support for protectionism is in line with their ideological profile, at least, at the party level: in the case of the populist left it is part of their economic redistribution agenda and their aversion to economic inequality, while in the case of the populist right, protectionism is mediated by their ethnocentric and nationalist cultural agenda. Even at the electorate level, this study found that support for protectionism among right-wing populist voters seems to be part of their cultural concerns. According to the authors, this finding suggests that the politicization of trade openness by the populist right could be more linked to the GAL-TAN dimension than to the economic redistributive agenda of the left.
Finally, it has also been argued that the ideological boundaries between populist right and left could be blurring as a consequence of the evolution that the populist right is experiencing in the economic dimension. Indeed, there is research that shows that right-wing populist parties are evolving to more centrist and even left-wing positions on economic issues (De Lange, 2007;Ivaldi, 2015). For example, the longitudinal analysis of Rovny and Polk (2020) confirms that on average, the RRPP have generally moved from the right side of the economic spectrum to the center, including some explicitly left-wing proposals in their electoral manifestos. However, the authors point out that "their overall stance on the economy continues to be ambiguous and contradictory. In order to keep and expand their economically heterogenous social coalition, radical right economic outlooks continue to do several things at once-they remain blurry" (p. 251). That is, this evolution to the economic center would be a consequence of a strategy of ambiguity to attract its potential voters which have heterogeneous economic interests: (ex)industrial workers, lower grade whitecollar workers, as well as small business owners. Thus, while the populist right holds clear stances on its primary issues (e.g., immigration), it is strategically blurring its positions on secondary issues, such as economy, in order to adjust its message to a broader audience. This finding is reinforced at the electoral level. The study shows that the likelihood to vote for rightwing populist parties is strongly predicted by their positions on immigration and moderately by their positions on social lifestyle, while their stances on redistribution have no statistical effect. In addition, the economic program of those right-wing populist parties who have evolved to more centrist or leftist positions tend to be welfare chauvinistic: "the view that access to welfare should be restricted to the 'deserving' natives" (Schumacher and van Kersbergen, 2016, p. 301). That is, they defend social and economic rights to some extent but only for the native people, excluding those who are not considered part of the nation. This means that this economic evolution would be also mediated by the ideological core of these parties, namely, nativism.
Therefore, this study aims to contribute to a better understanding of the nature of populism in Western Europe. In particular, this article aims to find out whether underlying ideology trumps populism in these types of political organizations or not. That is, if left-right differences prevail over the (anti)populist agenda of European political parties. In order to do that, this study will analyze the ideological positions and cohesiveness of populist parties in Western Europe. Considering the main findings of the research previously discussed, I expect the following: H.1. Left-right ideological orientation prevails over populism when it comes to explain the positions of populist parties in economic and cultural issues (e.g., redistribution, immigration, nationalism, etc. H.2.1. The distance between the positions of populist and non-populist parties toward the EU and free market is greater than that between right-wing and leftwing populist parties.

METHODOLOGY
To assess these hypotheses, two sources of data are analyzed: first, I explore the characteristics of European populist parties at the party or elite level using the 2019 Chapel Hill Expert Survey, a data set commonly used in this field of research (Bakker et al., 2020); and second, I also analyze this political phenomenon at the mass or electorate level using the 2018 Round of the European Social Survey. I use the 2019 Chapel Hill Expert Survey (CHES) to measure the ideological positions of populist parties across Western Europe. The selection of the political parties for this part of the analysis has followed the "anti-establishment/anti-elitism" scale from this expert survey. As noted by previous research, anti-elitism is a necessary and central element of populism (Mudde and Rovira Kaltwasser, 2013;March, 2017;Rooduijn and Akkerman, 2017). This variable measures the salience of antiestablishment and anti-elite rhetoric of each party on a scale where 0 means a complete absence of anti-establishment rhetoric and 10 means a complete presence. As shown in Figure 1, those parties that attack the elites (above the horizontal axis) tend to be located at the two extremes of the left-right ideological scale, while parties that are not characterized by this type of rhetoric (below the horizontal axis) are closer to the center. Indeed, the relationship between left-right ideology and populism is not linear but U-shaped, in line with previous research that conclude that radical parties on both the left and the right are inclined to employ a populist discourse in Western Europe (Rooduijn and Akkerman, 2017). In the case of Eastern European populism, is more biased to the right (Appendix 1): only a few political parties are characterized by strong anti-elitist rhetoric and have a leftist ideological orientation (e.g., the Slovenian Levica, the Communist Party of Bohemia and Moravia and the Latvian Russian Union). As the main goal of this study is to analyze the triumph of ideology over populism, a balanced sample of populist parties located at different points on the ideological scale is needed, reason why the study will focus on Western European populist parties.
With regard to the selection of populist parties, I have selected those political parties that have a score of 5 or more on the scale of anti-elitism. In total, 61 political parties from 20 countries have been selected. On average, these parties score 7.33 on the antielitism scale and 6.45 on the variable which measures whether "The people, not politicians, should make the most important decisions." With regard to the ideological orientation, 28 of 61 of these parties have been identified as left-wing (scored below or equal to 4 on a scale of 0 to 10, where 0 is extreme left and 10 is extreme right) while 27 of 61 were identified as rightwing (scored higher or equal to 6). Only 6 parties were labeled as centrist (between 4 and 6 on the ideological scale). The threshold of 5 on the anti-elitism scale allows for the inclusion of political parties that have been classified as populist in previous studies even if they show moderate levels of anti-elitism (e.g., the Danish People's Party or Vox). However, to ensure that this threshold does not affect the main findings of the study, the same analysis has been performed by taking a higher threshold (above or equal to 7 on the anti-elitism scale). The change in threshold does not affect the main findings as can be seen in Appendix 2.
To evaluate the hypotheses, the following variables have been analyzed. First, the variable "GALTAN" has been selected to measure the position of the parties in 2019 in terms of their views on social and cultural values, classifying the parties as "Libertarian/postmaterialist" or "Traditional/authoritarian" depending on whether the parties favor expanded personal freedoms or they reject them in favor of order, tradition and stability. Second, the variable of "LRECON" has been selected to measure the positions of political parties in terms of their ideological stance on economic issues (privatization, taxes, regulation, government spending, and the welfare state). With regard to the main issues addressed by Western European populist parties, the variables of IMMIGRATION_POLICY and NATIONALISM have been selected to assess the nativism of populist parties. As stated by Mudde (2007, p. 22), "the nativist dimension includes a combination of nationalism and xenophobia." In the case of left-wing populism, we include the variable of REDISTRIBUTION, a primary issue for leftwing populist parties (Fernández García, 2019). Finally, the overall orientation of the party leadership toward European integration in 2019 (EU_POSITION) as well as the position toward trade liberalization/protectionism (PROTECTIONISM) have been selected to assess the second hypothesis.
The analysis has proceeded as follows. First, the average positions held by the populist parties on the left and the right in the economic and cultural dimensions, as well as in the specific issues pointed out (immigration, EU, etc.), have been assessed to determine to what extent they tend to converge or differ. Second, the relative salience that the cultural and economic dimensions have for these parties has been examined to check whether they differ or not in the dimensions in which they mainly compete. Finally, I also analyze the ideological cohesiveness among populist parties on the right and the left in contrast with the cohesion among populist parties considered as a whole (without any ideological distinction) in order to evaluate whether ideology prevails over populism or not. In line with previous research (e.g., Camia and Caramani, 2012), I take the standard deviation of the positions held by the political parties in the different dimensions and issues mentioned as an indicator of ideological cohesion or dispersion 5 : The lower the value of the standard deviation, the more ideological homogeneity or cohesiveness within the type of populist party. To reinforce the results obtained through the previously mentioned analysis, the ideological cohesion of all Western European political parties (those included in the survey) and the distance between the positions of the different types of parties in the mentioned issues will also be analyzed. This will allow us to conclude whether the populist agenda of these actors tend to prevail over their left-right ideological orientation or not.
Finally, the main findings of the first analysis are contrasted at the electorate level in a selection of left, right and ideologically ambiguous populist parties using the 2018 Round of the European Social Survey. I compare the average ideological positions of right-wing and left-wing populist parties' voters in a selection of Western European countries that have witnessed the advance of both types of populism: the Netherlands (PVV and SP), France (RN and FI), Spain (Vox and Podemos) and Belgium (VB and PVdA-PTB). I also include the case of Italy, with the right-wing Lega and the ideologically ambiguous Five 5 The standard deviation measure varies from a minimum of 0 (when all the parties considered are in exactly the same ideological position on the different scales) to a maximum of 7. This is the maximum possible value of the SD calculated with the minimum number of possible cases (two parties) where one is located at one end of the scale (e.g. 0 -extreme left) and the other on the other extreme (e.g. 10 -extreme right). On the scale "Position toward the EU" ranging from 0 to 7, the maximum deviation would be 5. Stars Movement. The variables selected cover the main issues addressed in the first analysis: positions toward immigration (from an economic and ethno-racial perspective), social and cultural values ("Gay male and lesbian couples should have the same rights to adopt children as straight couples"), redistribution ("The government should take measures to reduce differences in income levels") and positions toward the European integration ("Unification has already gone too far-should go further"). The results obtained in this part will be also contrasted with a multivariate analysis that includes sociodemographic control variables, left-right ideological orientation as well as variables that may indicate a more anti-establishment profile (distrust of politicians and dissatisfaction with the functioning of democracy).

RESULTS
The first hypothesis stated that "Left-right ideological orientation prevails over populism when it comes to explain the positions of populist parties in economic and cultural issues (e.g., redistribution, immigration, nationalism, etc.)." Sub-hypotheses 1.1, 1.2, and 1.3 further establish that these parties differ in the dimension they compete (cultural, in the case of the populist right, and economic, in the case of the left) and that the ideological cohesion among right-wing and left-wing populist parties will be greater in their primary issues (nativism for rightwing populism and redistribution for left-wing populism) while it will be less so in secondary issues (redistribution for right-wing populism and nativism for left-wing populism). First, right-wing populist parties give more importance to the cultural dimension (GALTAN) than ones on the left (7.3 and 5.8, respectively, on a scale from 0 -not importantto 10 -very important-). The position of populist parties on the GALTAN scale (0: Libertarian/Postmaterialist -10: Traditional/Authoritarian) also maintains a positive linear relationship with the left-right ideological position 6 : right wing populist parties tend to be much more traditionalist and/or authoritarian (8.2 on average) than left wing populist parties 6 Pearson's R: 0.844, p < 0.01.
(2.8). Consequently, the cohesion among the populist right (SD: 1.5) and the populist left (SD: 1.6), as well as that of the parties that are located in the center (SD: 1.7) is considerably greater in this dimension than that of the populist parties considered as a whole (SD: 3) (see Table 1).
With regard to the primary issues of right-wing populism, namely, nativism, we can observe that the positions which populist parties have on immigration and nationalism also maintain a positive linear relationship with the left and right ideological orientation: right-wing populist parties tend to be much more nationalistic (8.8 on average) and anti-immigrant (9.2) than those on the left (3.3 and 2.4, respectively) (see Table 2 for mean values). The saliency of the migratory issue is also positively correlated with ideology: the more to the right, the more importance is given to the migration issue (Pearson's R: 0.836, p < 0.01) 7 . As shown in Figures 2, 3, right-wing populist parties, especially those located at the extreme right of the ideological scale (between 8 and 10), show high levels of cohesion on immigration and nationalism, while left-wing populist parties are more dispersed on these issues: they are generally liberal toward immigration and hold a cosmopolitan conception of society (e.g., the Spanish Podemos or the Portuguese Left Bloc) but some hold more moderate positions on immigration and have a more particularistic conception of society (e.g., the Irish Sinn Féin and the Dutch Socialist Party). As expected, the cohesion among the populist parties on the right (SD: 1.1 in immigration and 1.5 in nationalism) and left (SD: 1.2 in immigration and 1.7 in nationalism), as well as that of the parties that are located in the center (SD: 2.2 in immigration and 1.2 in nationalism), is considerably greater than that of the populist parties considered as a whole (3.5 in immigration and 3.1 in nationalism).
Second, economic issues are more important for left-wing populist parties (6.9, on a scale of 0 to 10, where 10 is "very important") than for right-wing populist parties (4.9). The ideological stance on economic issues maintains an almost perfect positive linear correlation with the left and right 7 The survey does not include a variable of salience for nationalism. ideological orientation 8 , being the cohesion among populist parties on the right (SD: 1.4), left (SD: 1.1) and center (SD: 0.8) much greater than that of the populist parties taken as a whole (SD: 2.7). With regard to redistribution, a primary issue for left-wing populism, we can observe that these parties are clearly positioned in favor of redistributing wealth (1.6 on average) while right-wing populist parties tend to be more opposed to it (6). The saliency of this issue is also correlated with ideology: the more to the left, the more visibility is given to the issue of redistribution (Pearson's R: −0.610, p < 0.01). As shown in Figure 4, left-wing populist parties display higher levels of cohesion on redistribution than on immigration and nationalism (Figures 2, 3) while right-wing populist parties are more dispersed on this issue: they move between positions very opposed to redistribution (e.g., the Spanish Vox), more 8 Pearson's R: 0.917, p < 0.01. moderate positions (e.g., the French National Rally) and others clearly favorable to it (e.g., the Greek Golden Dawn). Although the gap between right-wing and left-wing populist parties is less pronounced on redistribution than on immigration and nationalism, the cohesion among left (SD: 1) and right populist parties (SD: 1.6), as well as that of the parties that are located in the center (SD: 1.4) is still considerably greater than that of the populist parties considered as a whole (SD: 2.5).
With regard to Hypothesis 1.4, the analysis shows that the gap between right-wing and left-wing populist parties on cultural and economic issues is indeed wider than that between populist and non-populist parties (see Table 2). On the different issues analyzed, the distance between the left and right-wing populist parties exceeds 4 points. On the contrary, the distance between the positions of left-wing populist parties and left-wing nonpopulist parties only reaches 1 point on economic issues, while in the rest of the issues it remains below 1. The distance between  the positions of right-wing populist and non-right-wing populist parties is also very low on economic issues and redistribution (< 1), but is higher on immigration and the GALTAN dimension (> 2) and, especially, on nationalism (> 3). Therefore, there is a greater divergence between populist and non-populist parties on the right than on the left. Nevertheless, the gap on these cultural and economic issues remains wider between left-wing populists and right-wing populist parties than that between populists and non-populist parties.
The second hypothesis stated that "Populism prevails over left-right ideological criterion in the positions that these political parties hold on the European Union and protectionism." In Figure 5, we can observe that the position toward the European Union maintains a moderate-low linear correlation with the leftright ideological orientation: the sign of the correlation suggests that opposition to the EU is higher on the right (2.5 on average on the right, 4 on the left), although the intensity of the correlation is rather weak (0.4). The saliency of this issue is also higher in right-wing populist parties (6.5) than in left-wing populist parties (5.4), although the difference is less pronounced than in the economic and cultural dimensions. The cohesion shown by the right-wing populist parties (1.51) is practically the same as that of the left-wing parties (1.54); and, contrary to what is expected in hypothesis 2, the dispersion of the positions of the  parties considered as populists (1.7) is greater than if we consider them as left and right, although on this occasion the difference is fairly small.
With regard to protectionism, this variable does not maintain a linear correlation with ideological orientation. In Figure 6 we can observe that the position maintained toward free market has more to do with the degree of ideological radicalism than with the left-right ideological orientation: the closer to ideological extremes, the more protectionist the organizations are (e.g., the Greek KKE on the left, and the French National Rally on the right). On average, left-wing populist parties are slightly more protectionist (6.8) than those on the right (6.4), although the differences are very small. In this case, the degree of dispersion among the populist parties considered as a whole is less (SD: 1.8) than that of the right-wing populist parties (SD: 2), almost equal to that of the left (SD: 1.75) and greater than that of the center (SD: 1.6). By considering a higher threshold of anti-elitism to select the sample of populist parties (=>7 in the anti-elitism variable) (Appendix 2), the dispersion in these two issues -EU position and protectionism-is reduced in both the right and left parties, as well as those considered as populist without ideological distinction. This suggests that the more anti-establishment, the more cohesive the parties in European and protectionist matters are (they are also more Eurosceptic and protectionist). However, the cohesion among right-wing populist parties is in this occasion slightly higher than that of the populists considered as a whole, reason why we cannot fully confirm hypothesis 2 9 .
As for hypothesis 2.1, the analysis ( Table 2) shows that the gap between right-wing populist and left-wing populist parties on protectionism is smaller (below 1) than that between populist and non-populist parties. Again, the distance between populist and non-populist parties is greater on the right (3.1) than on the left (1.6). The same is true for the positions held toward the EU (3.1 and 1.5, respectively). In this case, however, the gap between right-wing and left-wing populist parties (1.5) is only slightly smaller than that between populists and nonpopulists on the left (1.52), thus hypothesis 2.1 can only be partially confirmed.
Consequently, the average cohesion of Western European parties is higher on the cultural and economic issues analyzed if we classify them by their left-right ideological orientation than if we classify them by their anti-elitist rhetoric as shown in Figure 7 (the smaller the dispersion values, the greater the cohesion of the parties in the different issues). On the contrary, the average cohesion of Western European parties with regard to the EU and the free market is higher if we classify them by their anti-elitism than if we do so by their positions on the left and right ideological scale. That is, Western European parties appear slightly more cohesive when classified as populist/nonpopulist with respect to the positions they hold toward the EU and free market, while they are better classified as left-right on the different cultural and economic issues. In Appendix 3, the 9 As pointed out in the methodological section, the consideration of a higher anti-elitism threshold does not change the rest of the findings. specific dispersion levels of Non-Populist/Populist parties as well as Left, Center and Right-wing parties are also shown 10 .
Finally, the analysis at the electorate level shows that the voters of the selected right-and left-wing populist parties also present strong differences in ideological terms, reinforcing the previous results. All left-wing populist voters are more tolerant of immigration, in both ethno-racial and economic terms than the right-wing populist voters (Appendix 4). The most liberal toward immigration is the electorate of Podemos, followed by the electorate of France Insoumise, while the most restrictive is the electorate of the French National Rally. In the cultural dimension, voters of left-wing populist parties are also more liberal, for example when asked if homosexual people should have the same rights as others to adopt children. Again, Podemos voters are the most liberal in the sample while Lega voters are the most conservative. When it comes to redistribution, all leftwing populist voters are more in favor of governments taking measures to reduce income gaps than right-wing populist voters, which tend to be more opposed to it. The individuals most in favor of redistribution are those who voted for France Insoumise, while the most opposed are the Dutch Party for Freedom's voters. Finally, voters of populist parties on the left are on average more favorable of the European Union than those on the right, confirming the result found at the party level. Also, in line with the previous analysis, Podemos' voters are the most favorable to the EU, while the most unfavorable are the Flemish Interest (VB) and National Rally voters. With respect to the electorate of the Five Stars Movement (M5S), it is located on average on the ideological center of the right-left scale (5); on the left in redistribution (1.72); between the right and the left in the 3 | Binary logistic regression models: voting for a populist party on the left (Podemos, FI, SP, PVDA-PTB) or ambiguous (M5S) compared to voting for a populist party on the right (Vox, RN, PVV, VB, Lega).

Podemos (1)
Vox (0) FI (1)   Variables with a statistically significant relationship are shown in gray, indicating the direction of the relationship.
attitudes toward the EU (on the right, Vox' electorate is the only one more pro-European than that of the M5S) and immigration (closer to the right than to the left); and is, along with the Lega's electorate, the most conservative when it comes to gays and lesbians' rights. All the differences pointed out are statistically significant according to the T-test for independent samples, with the exception of the variable on the EU and gay-lesbian's rights in the Belgian case 11 . When controlling for sociodemographic variables (Table 3), the voters of the left and right-wing populist parties do not substantially differ in their positions toward the EU and in those variables that indicate a more anti-establishment profile (only in the Dutch case, PS voters tend to be less distrustful of politicians than the PVV electorate), while they do in the left-right ideological position and on immigration (with the exception, again, of the Belgian case). Moreover, the Podemos electorate appears as more liberal with respect to the rights of homosexual people than Vox's electorate, while the electorate of France Insoumise appears as more favorable to redistributive policies than RN's voters. As for the Italian populist parties, they follow the patterns of that of the Spanish, French and Dutch populist parties, although the fit of its regression model is rather weak.

CONCLUSION
The objective of this research was to contribute to a better understanding of the nature of populism in Western Europe. In particular, this article aimed to find out whether underlying ideology trumps populism in these types of political organizations or not. That is, if the left-right differences continue to be hegemonic when it comes to explaining the positions that these parties hold on the main issues on which they compete.
In order to do that, this study has compared the ideological positions and cohesiveness of Western European populist parties. First, the analysis provided empirical evidence for the hypotheses 1 and 1.1-1.3: left-right ideological orientation prevails over populism when it comes to explain the positions that Western European populist parties have on the main economic and cultural issues. In general terms, left-wing populist parties tend to be more egalitarian in economic terms and liberal in the cultural dimension, while right-wing populists are more traditionalist, authoritarian and nativist in the cultural dimension, and more right-wing oriented in the economic dimension. These findings are also confirmed by the analysis at the electorate level. The cohesion among populist parties on the left and right is also greatest in their respective primary issues (redistribution on the left, nativism on the right). On the contrary, the dispersion among these parties is greater in those issues that are secondary for them (nativism on the left, redistribution on the right). In addition, the analysis shows that the gap between left and right-wing populist parties is wider than that between populist and non-populist parties on these issues. Therefore, the fact that some right-wing populist parties show centrist or even left positions in the economic dimension or that some left-wing populist parties are not as liberal as one would expect in the cultural dimension does not mean that these are defining features of these types of parties or that the leftright ideological criterion has lost effectiveness in distinguishing populist parties. These blurring positions could be interpreted as a strategy of ambiguity to attract potential voters that maintain diverse interests and positions in secondary issues for these parties, as argued in the study by Rovny and Polk (2020) on the evolution of right-wing populist parties in the economic dimension. Besides the strong differences in the positions held by populist parties on the left and the right on the different issues analyzed, the greater internal cohesion showed by these parties compared to the cohesion of the populist parties considered as a whole (without considering their ideological orientation) also reinforces the argument that the left-right ideological criterion continues to be more effective than populism in explaining the nature of these political parties. That is to say, the left-right differences remain hegemonic with regard to the positions they maintain, especially on their primary issues.
Second, the empirical analysis does not clearly confirm hypothesis 2. This hypothesis stated that populism prevails over left-right ideological criterion when it comes to explain the positions that these political parties hold on the European Union and protectionism. On the one hand, left-wing populist parties seem to be less opposed to the EU than right-wing partiesthe latter also confer more importance to this issue-, which has been confirmed by the analysis at the electoral level, although the multivariate analysis showed that these differences are not maintained when controlling for sociodemographic variables. On the other hand, the levels of protectionism seem to be more related to ideological radicalism and anti-elitism than to left or right ideological orientation: the more radical and anti-elitist, the more protectionist, the differences between the left and the right being very small. The relationship of these variables with ideological orientation is therefore weak (more so in the case of protectionist attitudes), in line with hypothesis 2. Yet, the analysis of the cohesion among the different types of populist parties does not allow us to fully confirm hypothesis 2: the cohesion of populist parties taken as a whole is no greater than that of populist parties on the left and right, although the differences are quite small on these two issues. Nevertheless, the fact that the dispersion shown by populist parties as a whole (without ideological distinctions) on protectionism and the EU is the lowest of the issues analyzed indicates that they do indeed bear some relation to the populist nature of these organizations. In addition, the analysis shows that the gap between populist and non-populist parties is wider than that between left and right populists in the positions they hold toward the free market and the EU -the difference is less pronounced on this last issue-. Thus, hypothesis 2 cannot be fully confirmed, but the analysis suggests that there is a close relationship between populism and opposition to the EU and protectionism.
In this sense, the opposition to the EU and to free trade agreements has an important anti-establishment component. In the case of the EU, it seems clearer: insofar as the European project is the result of the "consensus" of the political elites and has resulted in a complex and distant political organization, it is not surprising that the EU is part of the populist attacks (Taggart, 1998). As for opposition to free trade agreements, these are often the result of opaque and secret negotiations, which can fuel anti-establishment attacks by populist parties. However, we cannot ignore what different investigations such as those of Plaza-Colodro et al. (2018) and Van der Waal and De Koster (2018) point out. The attitudes toward the EU and protectionism of populist parties are considerably mediated by their thick ideology: in the case of the populist left it is part of their aversion to economic inequality and neo-liberal policies, while in the case of the populist right, it is mediated by their ethnocentric and nationalist cultural agenda.
Recapitulating, this research suggests that the left-right ideological orientation seems to remain hegemonic in explaining the nature of populist parties in Western Europe. As noted by previous research, however, some parties hold positions that seem to contradict their left-right positioning (e.g., right-wing populist parties that are in favor of redistribution policies or leftwing populist parties that hold more particularist conceptions of society). But these blurring positions cannot be overestimated and lead us to conclude that populism prevails over the underlying ideology of these organizations, or that populist parties are essentially "the same thing." This research, in line with other empirical studies (e.g., Otjes and Louwerse, 2015;March, 2017, Fernández-García andLuengo, 2018), suggests that underlying ideologies are what better explain the nature of these populist formations: populist parties on the right and the left in Western Europe seem to be first radical right and radical left rather than populists. This differentiation is also identifiable in the attitudes of their voters (e.g., Van Hauwaert and Van Kessel, 2018) and in the institutional behavior of these organizations (Otjes and Louwerse, 2015). In a similar vein, the coordination of these political parties at the European level on the basis of their left-right ideological orientation rather than their populist orientation reinforces the argument that the underlying ideology prevails over populism in Europe. Especially if we consider that Euroscepticism seems to be one of the issues that brings these political parties closer together, as we have been able to verify in this research.
Against this conclusion it can be argued that the 2015 populist coalition government in Greece between the leftist Syriza and the rightist ANEL is an evidence of the contrary, that is, that populism triumphs over ideology. However, the situation of exceptionality in which this government was formed -under strong pressure from European institutions for Greece to accept the harsh conditions of the economic bailouts-makes it not a good example. In this sense, the conditions under which this populist alliance was formed put the opposition to the European Union in the foreground (Rori, 2016) -one of the issues on which the different populist parties come closest-. Likewise, the populist coalition in Italy between the Lega and the Five Star Movement in 2018 seems to play in favor of the thesis that populism triumphs over ideology. However, the short duration of this government (1 year and 2 months) and the ideological ambiguity of M5S does not make this argument very strong. In this regard, M5S can be considered one of the few "pure" or "post-ideological" populist parties in Western Europe (e.g., Ivaldi et al., 2017), thus we could say that it is the exception to the rule. The tendency, so far, has been for populist parties to organize themselves in the European institutions on the basis of their left-right ideological orientation rather than trying to promote a populist and Eurosceptic coalition at the European level. At the national level, the participation of right-wing and left-wing populist parties (e.g., FPÖ, PVV, FrP, True Finns, Podemos, etc.) in governments led by center-right and center-left parties or parliamentary support for them (e.g., the Danish People's Party, the Left Bloc, etc.) are also becoming increasingly common. This trend is consistent with the findings of this research: the gap between right-wing and left-wing populist parties on cultural and economic issues is wider than that between populist and non-populist parties. This could explain why political alliances tend to be formed on the basis of the left-right ideological orientation rather than on the populist/anti-populist agenda of the parties. This situation could change, however, if opposition to the European Union or to denationalization processes in general becomes a predominant issue, hence widening the gap between parties opposed to these processes (populist parties) and parties in favor (center-left, center-right and liberal parties), as it happened in Greece in 2015.

AUTHOR CONTRIBUTIONS
BF-G conceptualized the research, conducted the analysis, drafted, and reviewed the manuscript.

FUNDING
This work was supported by Andalusian Plan for Research, Development and Innovation (PAIDI 2020).