1The 2017 French presidential election was notable for the unprecedented success of populist candidates and parties. On the right of the political spectrum was Marine Le Pen and, on the left, Jean-Luc Mélenchon. Beyond France, the electoral rise of populist parties and entrepreneurs is one of the most remarkable and heavily-studied political phenomena of the last two decades.  The proliferation of parties and movements that come under the umbrella of “populism” raises at least two major questions. The first is that of the unified nature of the phenomenon and our ability to distinguish its characteristic features. Are there commonalities between the many occurrences of populism manifested across the entire ideological spectrum?  In recent years, an impressive amount of international literature has tried to define this “plural” populism and identify its manifestations within party competition across different historical and political contexts. 
2A second significant area of investigation relates to the origins of populist mobilization in these diverse contexts, and particularly the adoption of populist ideas by voters. Recent work has stressed the existence of a populist “demand” and emphasized the presence of populism as a system of attitudes that guides individual electoral choices and motivates votes for populist candidates and parties. 
3As part of this line of research, the present article attempts to clarify the origins of the populist mobilization during the French presidential election of 2017. Based on a “minimal” definition of populism and its distinctive traits as a “thin-centered” ideology—with regards to its relation to the people, the elite and the expression of popular sovereignty—the article considers populism as a coherent set of attitudes among voters, whose impact can be differentiated from the more substantial ideologies and value systems it is based on, and from which it derives its dominant political connotation.
4Based on an analysis of two national surveys carried out during the 2017 elections, this article aims to empirically assess the effect of this populism “from below” on electoral choices in the first round of the presidential election. The paper asks the following questions: Is it is possible to identify populist attitudes among French voters in 2017 which shaped populist voting at either end of the electoral spectrum? Do these attitudes themselves have an effect? Finally, how do these populist attitudes interact with the main dimensions of socio-economic and cultural competition and more established ideological orientations, and particularly left and right ideological affiliations?
5The article begins by looking at the main defining features of populism, its interpretation as an attitude system, and the ways in which it overlaps with the more substantial ideologies on which it is based. We then try to demonstrate the existence of populist attitudes amongst French voters, and how they are distinct from other forms of political distrust or disengagement from electoral politics. Finally, the article attempts to demonstrate that populism has a distinct, substantial effect on voting, but that it does not challenge more traditional ideological identifications, for which populism plays a role of “supplementary” mobilization mechanism for individuals less pronounced preferences on the socio-economic and cultural dimensions of electoral competition.
The ideational approach to populism
6Populism has undoubtedly been one of the most discussed concepts in academia since the 1960s. In spite of its popularity in the media and academia, there is still no unified theory of populism.  Some scholars even consider it too heterogeneous to be categorized,  while others define it essentially as a style or political strategy.  This article does not attempt to bring the debate to a close. Rather, we point out that, in the last fifteen years, a relative consensus has been reached at the international level, centering on an “ideational” approach to populism as a discourse, ideology, or set of ideas.  Consequently, as Cas Mudde suggests, it is possible to identify the distinct features which make up a “minimal” definition.  This approach works at a sufficient level of abstraction to allow the concept of populism to be transposed to various national or regional contexts—such as the comparative analysis of populism in Latin America and Europe—and to identify the necessary and sufficient conditions for its empirical operationalization. 
At the heart of populism
7Approached as an “ideational” construct, populism can firstly be defined at least as a principle of organization and legitimization of power which centers on three core concepts: the people, the elite, and popular will.  As the political scientist Matthijs Rooduijn has emphasized, we should focus on the ideas that make up the ideational “core” of populism, and ignore certain elements which could be deemed more “peripheral,” like relations to leadership, organizational models, and demagogical style.  The idea of the people is the central focus of populism, and an appeal to the people lies at its core. Populists describe the people as an idealized symbolic community, a “pure” and homogeneous global political subject. This is built around a series of cultural, ethnic or socio-economic meanings.  In this sense, populism is above all an ideological monism: its vision functions according to a “logic of equivalence”  or “simplification”  through which it articulates a plurality of socio-political needs and demands from a variety of social groups, and constructs a common identity for them. The populist sees the people as systematically opposed to the elite. Populism is based on this polarizing, vertical representation of an essentially binary political society, divided into two antagonistic blocs, one subordinate (the people), the other dominant (the elite). The populist elite is interpreted in opposition to the people as a homogeneous, undifferentiated group of people, an “oligarchy” or “caste” defined by the possession of political, economic, cultural, or media power.
8Populism stresses the “equilibrium necessary between representatives and represented” and repudiates the distance between the elite and the people.  The opposition between the people and the elite is expressed in deeply Manichean terms of good versus evil. Populists perceive the elite as essentially “corrupt,” not just morally—through the denunciation of plutocracy and political racketeering—but even more deeply, because they claim that it ignores or betrays the true interests and will of the sovereign, “virtuous” people. The expression of the popular will is the third component of the ideational core of populism. This sees the sovereign people as the exclusive source of political power. Populism is based on the affirmation of the absolute unmediated supremacy of popular sovereignty, without obstacles or restrictions, and the idea of the “authenticity of the popular will liberated from the artifices of representation.”  According to populists, “people” possess will and “common sense,” but have been dispossessed of their sovereignty by an elite that ignores their demands. Inevitably, populist leaders claim to have a monopoly over the representation of this “authentic” people, claiming to speak for the “forgotten,” the “people,” the “silent majority.” This claim to act as the sole representative of the “subordinate” and the “excluded” is a central feature of populism, which is characterized by its rejection of consensus and political compromise and harsh criticism of the principles of representative democracy. At the same time, populism is not antithetical to democracy, and constitutes what Margaret Canovan refers to as a “redemptive” approach.  Yves Mény and Yves Surel see populism as a rebalancing force within modern democracy which favors its “popular element.” Populism promotes and reinforces the democratic ideal through its vision of the exclusive power of the demos.  Because it defends popular sovereignty and claims to give power back to social groups who no longer feel represented by traditional parties, populism can present itself as a “democratizing” and “re-politicizing” force for issues that have been abandoned by the major governing parties. 
Populism as a “thin-centered” ideology
9A second idea within the ideational approach is that of populism as a “thin-centered” ideology —that is, relatively impoverished because it does not propose a comprehensive vision of the world, unlike more densely structured ideologies like Marxism or liberalism.  As a “thin-centered” ideology, populism is based on thicker ideologies which give it substance and provide it with the concepts necessary to develop its program of action and mobilization. Contemporary research on populism defines two main manifestations at the macro level, one on the left and the other on the right.  This definition is based on a multi-dimensional representation of the electoral political space, structured by two dimensions of competition—“materialist” socio-economic on the one hand, and cultural on the other—in line with the traditional political identities of right and left .  On the right of the spectrum, populism is most often found within the radical right. It may be defined as culturally exclusive, and so it operates mainly on the non-material dimension of competition. This populism politicizes issues connected to immigration, to authority, and to the national identity of its “people,” who are idealized and primarily imagined in cultural terms as an ethnos.  The Front National in France is typically considered the prototypical model of radical right populism in Europe.  On the left, populism is associated with socio-economic issues and opposition to economic liberalism, modeled especially on popular movements in Latin America at the turn of the twenty-first century. This is an inclusive populism, underpinned by egalitarian, humanistic, universal values—claiming to speak for disadvantaged social groups and minorities, while stressing their opposition to a common “other,” represented by the political establishment and/or the neoliberal elite.  In France, the trajectory of Mélenchon and La France Insoumise (LFI) demonstrates the crystallization of left-wing populism in the political arena, a vox populi distinct from Marxism and the international workers movement which has particularly emerged from the dialogue between Mélenchon and the leading figures of post-Marxism, Chantal Mouffe and Ernesto Laclau.  Finally, different types of populism interact with a third aspect of competition, related to internationalization and European integration. This dividing line, which in France has been described by Pascal Perrineau as an opposition between an “open” and a “closed society,”  overlaps with both the socio-economic and cultural dimensions. Hanspeter Kriesi and his colleagues argue that political competition now operates between two poles of “demarcation” and “integration,” pitting globalization’s “losers” against its “winners.”  Initially, these authors saw the parties of the radical right as the main actors of the demarcation pole. More recently, Kriesi and Takis Pappas have highlighted the impact of the “great” economic recession. In their view, the context created by the crisis has allowed the populist left, grounded in socio-economic antagonism, to mobilize issues of denationalization.  Enrique Hernández and Kriesi demonstrate that the economic crisis created a favorable framework of opportunity for a populist, national-protectionist left.  This resistance by populism to international openness is expressed through opposition to Europe. Euroskepticism is a feature shared by radical parties of the left and right.  Marianne Kneur has recently described opposition to the European Union as a common feature of populist parties, seeing this new “tandem” as the confirmation of a transnational divide along which populism is aligned, which is independent of the left-right cleavage. 
A populism “from below”: Populist attitudes and opinions
10The third and final aspect of the ideational approach is that it allows populism in voters to be considered objectively. Recent empirical work has demonstrated the existence of populist attitudes in the population, which are distributed unequally among social groups, are spread across the whole political spectrum, and can be shown to exert normative influence on electoral choices.  The empirical analyses in this article are based on the instrument for the measure of populist attitudes, as implemented by Agnes Akkerman, Cas Mudde, and Andrej Zaslove. This scale is based on the “minimal” definition of populism proposed above, and includes three pairs of items, which aim to empirically distinguish the characteristic traits of populism (Table 1). 
Items on the populism scale*
|POP 1||Politicians of the Assemblée Nationale need to follow the will of the people.|
|POP 2||The people, and not politicians, should make our most important policy decisions.|
|POP 3||The political differences between the elite and the people are larger than the differences among the people.|
|POP4||I would rather be represented by a citizen than by a specialized politician.|
|POP5||Elected officials talk too much and take too little action.|
|POP6||What people call “compromise” in politics is really just selling out on one’s principles.|
Items on the populism scale** Ordinal scales from 1 (strongly disagree) to 5 (strongly agree).
11POP1 and POP2 reflect adherence to the principle of an exclusive and unmediated supremacy of popular sovereignty, which sees the people as the only legitimate source of power, free from the constraints of representation. It should be noted that these two statements express the “democratic” ideal of populism, its claim of being “redemptive,” to use Margaret Canovan’s term. Taken in isolation, these statements can be considered as markers of attachment to democratic principles.
12In order to be populist, appeals to popular sovereignty must be combined with a Manichean vision of a disconnection between the people and the elite and criticism of representation. POP3 measures the idea of a distinction between people and the elite, representing the perceived distance between the two groups. POP4 evokes the negative view of representative government. Finally, POP5 and POP6 are related to negative Manichean judgments of political elites, which make up the final aspect of the populist ideology.
13The scale satisfies two important requirements. Firstly, it attempts to operationalize the distinctive features which make up the ideological “core” of populism. Here it is crucial to emphasize that the six statements that make up the scale cannot be separated from one another: it is their simultaneous presence that identifies populism as a system of representation and of organizing power, and differentiates it from other political attitudes like anti-establishment attitudes, political distrust, and disengagement.  Secondly, the scale allows us to explore interactions between populism as a “thin-centered” ideology and the more substantial values and political identities which traditionally shape electoral choice.
Populist attitudes and opinions in the French presidential election of 2017
14How does this apply to populist ideas in the French electorate in 2017? To answer this question, the above “populism” items were submitted to respondents to two national electoral surveys conducted during the 2017 presidential election: the SCoRE survey, carried out as part of the European research project Sub-national Context and Radical Right Support in Europe, and Cevipof’s Enquête Électorale Française (ENEF) (see details of these two surveys in the methodological appendix.)
15The responses to the items attest to the strength of populist ideas during the election (Figures 1a and 1b). Not surprisingly, the affirmation of the supremacy of popular sovereignty (POP1 and POP2) gets high average scores on the scale of 1 to 5: 4.4 and 4.1 in the SCoRE survey, and 4.3 and 3.8 in the ENEF survey. Most respondents perceived a real gap between people and elites, with average scores for POP3 and POP4 of 4.2 and 4.1 (SCoRE survey) and 4.1 and 3.6 (ENEF survey), respectively. Finally, the extent of political distrust in France was revealed once more: a very large majority felt that “politicians talk too much and do not act enough,” as shown by the average score of 4.4 for POP5 in both surveys. Finally, and to a lesser extent, political compromise is viewed by many as being “a renunciation of one’s principles” (receiving average scores of 3.6 and 3.4 for POP6 in the SCoRE and ENEF surveys).
Data: SCoRE 2017 election survey, May-June 2017 (N = 16,546); ENEF-Cevipof survey, wave thirteen, 16–17 April 2017 (N = 8,119).
ENEF survey wave thirteen
ENEF survey wave thirteen
17These data demonstrate that a large portion of the French electorate is receptive to the rhetoric of populist parties. The levels observed in France are comparable to those identified from the same set of questions in several other European countries, which places France alongside other southern countries like Italy, Spain, and Greece, where populism is most developed.  Taken individually, however, each item represents only one facet of populism, a component of its ideational “core.” Empirically, the scale refers to a “latent” dimension of populism, or coherent system of attitudes, from the individual items. The homogeneity of the responses is verified by a Cronbach’s alpha of 0.77 for each of the two surveys.  The hypothesis of a “latent trait” of populism is validated further by a Mokken analysis, which enables a more restrictive set of tests than the alpha alone.  In each of the two surveys, the homogeneity coefficients H of the six individual populism items are all above the generally accepted threshold of 0.3, confirming the unidimensional nature of the scale. The H scalability coefficients of the two scales are above 0.4—rule of thumb is H > 0.4. These two surveys thus represent a sufficiently robust “scale” of populism. A unique populism “score” is calculated for each respondent from the average of the responses to each statement on the scale. The distribution of the respondents across the index obtained is presented in Figure 2. The average populism scores are 4.1 (SCoRE survey) and 3.9 (ENEF survey) on a scale of 1 to 5.
Data: SCoRE 2017 election survey, May-June 2017 (N = 16,546); ENEF-Cevipof survey, wave thirteen, 16–17 April 2017 (N = 8,119).
19The distribution of respondents on the populism scale in these two samples shows the extent of the phenomenon in the French political space during the 2017 elections (Figure 2). We note more pronounced asymmetry in the SCoRE data, with a negative skew (skewness = -0.9). This populist “bias” reflects the initial distributions of individual items. The distribution of populist attitudes is less strongly skewed to the left in the ENEF survey (skewness = -0.4).
20While they are widely shared, populist attitudes were nevertheless not distributed in the same way across different electorates. Figure 3 compares the average level of populism in different electorates for each of the two surveys. We observe that populism clearly differentiates the supporters of the three mainstream candidates—Benoît Hamon, Emmanuel Macron, and François Fillon—from those who voted for the populist candidates—Marine Le Pen, Jean Luc Mélenchon, and Nicolas Dupont-Aignan—whose average scores were significantly higher than typical for the French electorate in 2017. In both surveys, Front National voters showed the highest levels of populism of the entire electorate.
* Average “populism” scores observed for each group. The vertical bars represent 95% confidence intervals (CIs) around the mean. Base = valid answers, individuals who declared a vote in the first round. Others = Jean Lassalle, Philippe Poutou, French Asselineau, Nathalie Arthaud, and Jacques Cheminade. Data: SCoRE 2017 election survey, May-June 2017 (N = 13,780); ENEF-Cevipof survey, wave thirteen, 16–17 April 2017 (N = 7,091).
22These differences are all statistically significant, and very similar distributions can be observed in each of the two surveys, with the notable exception of Fillon voters, who demonstrated the lowest levels of populism in the ENEF survey. This difference is partly due to the under-representation of Fillon voters in the SCoRE survey, particularly older voters, who, as we will see, are statistically less inclined towards populism.
Populism as a distinct normative dimension
23The relationship between populist attitudes and electoral choices in 2017 can of course conceal more complex effects of sociological and attitudinal factors. Furthermore, we must grasp the influence of populism itself on the vote in order to identify its distinct normative effect. The impact of populist attitudes is tested through three series of binary logistic regression models.  For each series, votes for Mélenchon on one side and Le Pen on the other are contrasted with votes in favor of the three leading mainstream candidates, Fillon, Hamon, and Macron. These different models are nested in order to verify the validity and robustness of the effects observed for different specifications. In models 1a and 1b, the effect of populist attitudes is first examined controlling for standard socio-demographic factors of gender, age, level of studies, profession, religion, and assets; models 2a and 2b introduce a set of cognitive and behavioral correlations of populism; finally, models 3a and 3b add the left-right ideological affiliations of respondents as well as their positions on the main dimensions of competition. The results of these different models are presented in Table A1 in the methodological appendix.
Sociology of the populist vote
24Models 1a and 1b confirm the socio-cultural characteristics of the populist vote outlined in the comparative literature.  Votes for Mélenchon and Le Pen follow the same logic of age and education: these two groups of voters can be clearly distinguished from supporters of mainstream parties by virtue of their youth and weaker cultural capital—particularly Le Pen voters, among whom the effect of education is most noticeable. The data confirm the rooting of the populist vote in the lower social strata, especially among employees and workers. In addition, there is a strong negative correlation between the probability of voting for Mélenchon or Le Pen and assets, which partly reflects the divide between the wealthy and the non-wealthy in France.  However, Mélenchon voters appear more “cross-class” in character  while Le Pen’s electorate is strongly anchored in the working class, an observation in line with what is also known about the proletarianization of the Front National (FN) vote  Models 1a and 1b demonstrate the specific effect of populist attitudes: the probability of voting for Mélenchon on the one hand and Le Pen on the other is positively correlated with voters’ populism, regardless of their socio-cultural characteristics. In both cases, this effect is substantial and statistically significant, and warrants further assessment.
Populism, economic grievances, and political alienation
25Models 2a and 2b enable us to examine the distinctiveness of populism with regards to other mechanisms traditionally used to explain electoral support for peripheral protest parties like LFI and the FN. First, populism, as Benjamin Moffit in particular suggests, is closely linked to the idea of “crisis.”  Votes for these parties can be considered a political expression of economic grievances, a feeling of subjective poverty fueled by the crisis.  Second, the populist vote can be the result of a lack of political efficacy which is both “internal” (the impression of political competence and that one can have an effect on the political system) and external (trust in the ability of the system to respond to one’s needs and expectations.)  Finally, populism can represent part of a larger process of political alienation, characterized by indifference to politics and decline of partisan attachments.  These various mechanisms are tested alongside populism in models 2a and 2b.  Two of these variables have a significant, transversal effect on the electorates of Mélenchon and Le Pen. On the one hand, the probability of voting for each of the two populist candidates increases significantly with subjective feelings of economic hardship by individuals. On the other, external political efficacy—that is, the perceived ability to exert influence on the government’s choices—significantly reduces the probability to vote for Le Pen or Mélenchon. While correlated with each of these two predictors,  populism nevertheless continues to have its own effect here. The positive coefficients show that an increase in populism resulted in a significant increase in the probability of voting for Mélenchon or Le Pen in the first round of the 2017 presidential election. Figure 4 presents the average marginal effects of populism and the main indicators in models 2a and 2b.  If we consider the average effect here as a linear approximation, a one-point increase on the populism scale produces an average increase of about 12% in the predicted probability of voting for Mélenchon (see Figure 4a) and 16% in the predicted probability of voting for Le Pen (see Figure 4b). More realistically, and given the observed distribution of populism scores, a 0.5 point increase in populism increases the predicted probability of a populist vote by about 6% for Mélenchon and 8% for Le Pen. Populist voters are inclined to prefer populist candidates regardless of the respondents’ subjective perception of their economic situation, their interest in politics and their sense of political competence and efficacy, independent of party identification, while also controlling for the effect of standard socio-demographic variables. 
Socio-demographic controls were not included in the graphs for greater clarity.
Data: SCoRE electoral survey 2017, May-June 2017
27These results confirm that populism is a distinct attitude, one that exerts a substantial normative effect on electoral choice and should not be confused with the expression of economic grievances, political skepticism, or political alienation. Furthermore, models 2a and 2b can be replicated by contrasting the populist vote for Mélenchon and Le Pen with abstention. For the first round of the 2017 presidential election, in both cases populist attitudes increase the theoretical probability of voting for each of the two populist candidates rather than abstaining.  The average marginal effects are weaker but remain significant: 3% for the Mélenchon vote (N = 4,037), and 7% for the Le Pen vote (N = 3,857). These results should be taken with some caution, however, given the traditional under-representation in opinion polls of relatively depoliticized social groups and the under-reporting of abstentionist behaviors.  The idea of populism as a “redemptive” democratic force, as described by Margaret Canovan, is nevertheless present here. Notwithstanding its clearly anti-system features, populism can serve as a “democratic corrective” and a vector for mobilizing electoral sectors disconnected from the large governing parties. In this respect, populism is doubtless more closely related to (re)politicization than disengagement, as the presence of a populist party seems to make it possible for individuals who would otherwise turn away from the political system to enter it, particularly those from lower-income social groups with lower levels of education. 
Populism, left and right affiliation, and dimensions of competition
28As a “thin-centered” ideology, populism attaches itself to more substantial ideologies that provide it with a fuller program of action. What about interactions between attitude-based populism and voters’ political identity and values in their voting choices?
29In this third section, we test the effects of populism on voting when “thicker” ideological orientations are added to the previous models. Two aspects are considered: on the one hand, the left-right affiliation of the voters, which we know to be a determinant factor for voting in France,  and on the other hand the main dimensions of socio-economic and cultural competition which represent a strong basis for political alignments.  For the economic axis, we use the preferences of voters with regards to redistribution of wealth. This dimension was identified as a major factor in voting for radical left-wing populism in Europe.  The cultural axis is more heterogenous. We suggest to break it down into different attitudinal dimensions, which essentially refer to the main structuring poles on the right-hand side of the GAL/TAN axis as defined by Liesbet Hooghe and her colleagues: traditionalism, authoritarianism, and nationalism. Traditionalism expresses adherence to traditional moral values. Authoritarianism refers to a vision of society based on order and authority.  Regarding nationalism, two aspects should to be distinguished: on the one hand, opposition to immigration; on the other, a more general attitude of “closure” to internationalization, expressed through resistance to European integration and economic globalization.  To summarize, in addition to all the previous socio-demographic and political controls, the effect of populism is tested alongside the left-right affiliation of the respondents and their positions on the five aforementioned attitudinal dimensions: redistribution, anti-immigration, authoritarianism, traditionalism, and attitude towards Europe and globalization (see details of the scales in Table 2). 
Items in the attitude scales and descriptive statistics
|Redistribution||Social benefits and services in France place too great a strain on the economy (REV)|
The government should increase taxes a lot and spend much more on social benefits and services
|Anti-immigration||Cultural life is generally enriched by people coming to live here from other countries (REV)|
It is generally good for France’s economy that people come to live here from other countries
Thinking about taxes and welfare, people who come to live here from other countries generally put in more than they take out (REV)
|Authoritarianism||What our country really needs instead of more ‘‘civil rights’’ is a good stiff dose of law and order.|
What our country needs most is disciplined citizens, following national leaders in unity
|Traditionalism||Homosexuals and feminists should be praised for being brave enough to defy ‘‘traditional family values’ (REV)|
It’s a good thing that same-sex marriage is equal to opposite-sex marriage in the eyes of the law (REV)
|Europe, globalization||European unification should go further|
Globalization is an opportunity for economic growth in France
Globalization increases social inequality (REV)
Items in the attitude scales and descriptive statisticsOrdinal scales from 1 (strongly disagree) to 7 (strongly agree).
REV = categories of responses reversed to move in the direction of the scale.
Data: SCoRE 2017 election survey, May-June 2017 (N = 12,956) (Validated response base).
30The results of models 3a and 3b are presented in Table A1 of the Methodological Appendix. The average marginal effects of the main factors are shown in Figure 5.
The socio-demographic and attitudinal variable controls of the previous models were not included in the graph for readability reasons.
Data: SCoRE 2017 electoral survey, May-June 2017.
32These results confirm the specific effect of populism on voting for the two main populist candidates in the first round of the 2017 presidential election. A one-point increase in the populism scale produces an average increase of 8% and 6% respectively in the theoretical voting probabilities for Mélenchon (Figure 5a) and Le Pen (Figure 5b).  This effect remains positive and significant, showing the distinct normative effect of populist attitudes alongside more substantial political affiliations and ideological preferences.
33The significant marginal effects of voter left-right affiliation support the notion that populism is a “thin-centered” ideology rooted in more dense value systems, clearly differentiating left populism (Mélenchon) from right populism (Le Pen). The theoretical probability of voting for Mélenchon is strongly indexed to the position of the respondents on the left-right scale, and increases according to proximity to the left of the axis (Figure 5a). Conversely, a position on the right of the scale has a positive and significant effect on the likelihood of voting for Le Pen, demonstrating the largely right-leaning inclination of Lepenist populism (Figure 5b).
34The results that can be observed from the value dimensions also confirm, with slight variations, the idea of a distribution of populism along the two poles of an economically inclusive left, on the one hand, and a culturally exclusive right, on the other hand.  Not surprisingly, alongside populism on the right we find that opposition to immigration and authoritarianism—which are traditional attitudinal determinants of the Le Pen vote—have positive and significant average marginal effects (Figure 5b). On the left, the probability of voting for Mélenchon is positively correlated with support for more social justice and redistribution—which have dominated the LFI candidate’s campaign in 2017. A negative correlation with authoritarianism can also be observed (Figure 5a).
35While clearly different from the xenophobia of the Lepenists, Mélenchon’s support is however not distinct from mainstream voters with regards to attitudes towards immigrants. These results contradict recent comparative works which have observed a significant negative correlation between anti-immigrant prejudice and voting for populist movements of the left.  This divergence could partly be explained by the changes in Mélenchon’s positions, and his increasingly ambiguous stance towards refugees and immigration during the 2017 campaign.  Lastly, the results of models 3a and 3b suggest convergence of French populist movements on transnational issues. The rejection of European integration and the perception of globalization as a menace set populist voters on both sides of the spectrum apart from their counterparts in the parties of government. The average marginal effects of the two factors are significant and they are identical for Mélenchon voters (figure 5a) and Le Pen voters (figure 5b), who are both on the “closed” pole of this axis, in opposition to mainstream candidates.  Let us make two qualifications here, however. On the one hand, the literature suggests that the motives behind opposition to globalization may vary according to voters’ political affiliation: for right-wing populism, these are mostly immigration-related cultural motives; for left-wing populism, they are mostly economic reasons involving resistance to “neoliberal” hegemony.  On the other hand, Lepenist voters remain significantly more “closed” than their left-wing counterparts: a model contrasting populist voters (model 3c, Table A1 of the methodological appendix) shows that populism remains the only attitude common to both populist electorates (coeff. = -0.11 n.s.). They on the other hand diverge significantly on all the other attitudinal variables, including the transnational dimension, where Le Pen’s supporters remain more closed and show greater opposition to globalization and EU integration.
Is populism a “supplementary” mechanism?
36Finally, the effect of populism interacts with its attaching “host” ideologies, or more precisely, the size of the effect of populism may vary according to how far ideological preferences are ingrained. Hypothetically, populism should play a greater role amongst voters with ‘weaker’ political affiliations and/or less extreme positions on the main dimensions of competition. The two previous models are tested here by adding interaction effects between populism, left-right political affiliation and the main issue dimensions, namely redistribution for left-wing populism and anti-immigration for right-wing populism. These models are presented in Table A2 in the appendix.
37Overall, interaction coefficients are not significant, and each variable exerts its effect independently of the others. For the Mélenchon vote, the effect of populism is of the same magnitude regardless of voter support for redistribution or position on the left-right scale. With regards to the Le Pen vote, the interaction between populism and position on the left-right axis is not statistically significant. On the other hand, a significant effect (interaction coefficient -0,11, p < 0,05) can be observed for the interaction between populism and anti-immigration attitudes. The sign of the coefficient suggests that the effect of populism decreases as opposition to immigration increases, which supports the initial hypothesis.
38To delve further into the assessment of these interactions, we can divide the sample in two using the median of the populism scale, which allows the individuals to be distributed into two classes of populist attitudes: “weak” and “strong.” For each group, the probability of voting for a populist candidate is calculated from the logistic regression models, according to the main dimension of competition, and it is plotted together with a 95% confidence interval. The mean probability of populist voting is indicated by a horizontal line and serves as a reference to evaluate the different impact of populism across different locations of each issue dimension (see Figure 6). 
Populism classes around the median: Mélenchon vote (low N = 3,586, high N = 3,375); Le Pen vote (low N = 3,872, high N = 3,054).
Data: SCoRE 2017 electoral survey, May-June 2017.
40A second dynamic emerges from the graphs in Figure 6: populism may function as a “supplementary” mechanism for electoral mobilization. This mechanism responds to a spatial logic. Voters at the extremes of the issue dimensions appear motivated primarily by their ideological preferences: the propensity to vote for populist candidates is always significantly different from the average probability; populism in these cases simply reinforces the effect of left-right political affiliations and socio-economic or cultural values.
41The impact of populism is however more evident around median positions, at the center of the main dimensions of competition, among voters whose ideological preferences are less clearly defined. Populism plays a role as an additional driver of mobilization, somehow supplementing the weakness of political affiliations and ideological preferences among these specific segments of the electorate.
42With regards to the Mélenchon vote, this supplementary effect remains relatively limited for left-right affiliation. Populism operates here on a very small sector of the electorate, located in the center-left (around position 5, Figure 6a2). For these individuals only, strong populism increases the predicted probabilities of voting for Mélenchon, above the average probability of the sample, symbolized by the horizontal line. Elsewhere, voter left-right affiliation strongly determines the pro-Mélenchon vote (left) and the anti-Mélenchon vote (right), regardless of the degree of populism.
43There is a much clearer effect on the redistribution axis. Higher populism contributes significantly to an increase in the predicted probability of voting for Mélenchon in the median zone (positions 3 to 5, Figure 6a1)—that is, among an electorate less shaped by its preferences on the economic dimension of competition. Here, populism effectively allows an “over-mobilization” of these voters whose probability of voting for Mélenchon becomes higher than average when the level of populism is higher.
44For the Le Pen vote, attitudes towards immigration remain a determining factor. The effect of populism remains relatively marginal in this case, particularly with regards to more “cosmopolitan” individuals for whom the likelihood of voting for Le Pen remains far lower than average. Populism only works on a reduced section of the scale (around the 5 mark, figure 6b1), among voters moderately opposed to immigration, where we see a significant increase in the likelihood of voting for Le Pen for a higher level of populism.
45Conversely, the interaction of populism with left-right affiliation suggests a potentially greater impact of the degree of populism to the right of the political spectrum. The opposition between left and right is very clear here: regardless of their level of populism, individuals located on the left on the axis (position 5 and below, Figure 6b2) have a lower than average probability of voting for Le Pen. To the right of the axis, however, populism has a significant influence across most positions. With the exception of the most extreme category, a rise in the level of populism produces a significant increase in the predicted probability of voting for Le Pen, reflecting the importance of populism as a mobilization mechanism within the right-wing electorate. This effect is particularly visible in the central position of the left-right axis, where the degree of populism almost triples (from 0.12 to 0.35) the probability of choosing Le Pen in the first round of the 2017 presidential election—probably illustrating the presence, alongside moderately populist centrist voters, of a niniste group –i.e. voters who see themselves as neither left, nor right–, which as we know makes up a significant portion of Lepenist populism. 
46* * *
47Using an ideational definition of populism, this article has tried to empirically test the existence of populist attitudes within the French electorate in 2017, and the distribution of these attitudes along the political axis, hypothesizing that populism exists across the political spectrum, and particularly at its two extremes.
48The results demonstrate the existence of populism as a coherent set of ideas, organized by a monistic, polarized vision of a dual political society, the notion of a deep antagonism between “the people” and “the elite,” and the affirmation of the absolute supremacy of popular sovereignty. For voters, populism represents a measurable set of attitudes, oriented towards a class of socio-political objects related to power and representation, of variable intensity, which has an independent and significant normative effect on electoral behavior and orientation.
49Populist attitudes significantly increase the likelihood of voting for the two main populist candidates in the first round of the 2017 presidential election. Populism also distinguishes the vote for Mélenchon and Le Pen from the choice to abstain, and thus seems to constitute a vector for mobilizing voters who are disconnected from traditional parties. In this respect, it probably corresponds more to a logic of politicization than apathy or disengagement from politics.
50However, the determining influence of populism on electoral choice is not exercised in isolation. Our results confirm that, as a thin-centered ideology, populism needs attach itself to more substantial sets of ideas, and that it interacts with those to structure electoral behavior. The logic at work is additive rather than substitutive: populism reinforces the main effects of the thicker ideological forms. In particular, the distinction between left and right does not become “secondary.”  On the contrary, voter populism only takes on meaning when it is based on the values and ideological preferences to which it associates itself. A left-wing political identity provides the essential ideological foundation for the electorate of Mélenchon. On the right, Lepenist populism focuses primarily on cultural issues, primarily the rejection of immigration.
51Another dynamic is emerging, however: populism as a “supplementary” mechanism for electoral mobilization. This effect is spatially distributed. Populism specifically operates on the median positions of the main dimensions of competition, with voters whose ideological preferences are less pronounced and less extreme. Among these voters, populism acts as an additional driver of mobilization in favor of populist candidates, which can, if necessary, take some weight off their weak affiliations and ideological preferences. These conclusions are in line with those of Steven Van Hauwaert and Stijn Van Kessel, who see populism as a “motivational substitute.” 
52Finally, our results attest to the importance of the transnational divide in the mobilization of populist actors at both poles of the partisan spectrum during the 2017 presidential election.  Criticism of the European Union and globalization contribute strongly to differentiating populist voters from their mainstream counterparts. On the right, ethno-cultural nationalism traditionally is at the core of the FN’s ideology, similar to other radical right parties in Europe. On the left, based on the experience of Latin American national-popular revolutions, populism has recently theorized the need to “hegemonize national identification,” through the formulation of a “progressive left-wing patriotism” emphasizing social justice and equality.  For Mélenchon, the fight against neoliberal hegemony also asserts itself as a struggle for national sovereignty, an entry into resistance to the European Union and to globalization.
53The increasing intersection of populism with claims to national sovereignty is a fundamental element of contemporary populist dynamics,  as illustrated by the success of populist idiosyncrasy embodied by Donald Trump in the American presidential election of 2016.  This link reflects the tension that exists today between traditional political-ideological affiliations such as left and right in France, and the claim by populism to embody a “transversal” people. In any case, our results refute the idea that populism can be distinguished from more complex ideologies like left-right identity or value dimensions. Future research should examine how populism increasingly interacts with the transnational dimension of competition.
54The analyses presented in this article are based on data from two surveys conducted during the 2017 elections, based on representative samples of the French population.
55The SCoRE survey was conducted as part of the European project Sub-national Context and Radical Right Support in Europe. It was carried out online on behalf of the University of Nice by the BVA Institute, from 11 May to 25 June 2017. The sample contained 19,454 people, representative of the French population aged 18 and over in metropolitan France. Among these, 18,418 were registered to vote. The representativeness of the sample was ensured by the quota method and an adjustment applied to the following variables: gender, age, education level, region, and size of agglomeration.
56The Enquête Électorale Française (ENEF) by Cevipof-Sciences Po Paris is a national panel survey conducted since November 2015.  The data used in this article comes from wave thirteen, conducted on April 16–17, 2017 with a sample of 11,601 people, all of them registered to vote, constituting a national sample representative of eligible French voters. The sample was surveyed online and selected according to the quota method: gender, age, occupation of the reference person in the household, region, and category of urban area.
Logistic regressions, models with interactions
|Mélenchon versus mainstream||Mélenchon versus mainstream|
|Populism||0.62 (0.16)***||0.44 (0.12)***|
|Left-right placement||-0.36 (0.02)***||-0.54 (0.10)***|
|Redistribution||0.16 (0.15)||0.16 (0.03)***|
|Anti-immigration||-0.01 (0.03)||-0.01 (0.03)|
|Authoritarianism||-0.13 (0.03)***||-0.12 (0.03)***|
|Traditionalism||-0.04 (0.02)||-0.04 (0.02)|
|Europe, globalization||-0.56 (0.03)***||-0.56 (0.03)***|
|Intercept||3.40 (0.78)***||4.21 (0.67)***|
|N||6 961||6 961|
|Log Likelihood||-2 734||-2 732|
|Le Pen versus mainstream||Le Pen versus mainstream|
|Populism||1.19 (0.26)***||0.70 (0.2)**|
|Left-right placement||0.22 (0.02)***||0.29 (0.12)*|
|Redistribution||-0.10 (0.03)**||-0.10 (0.03)**|
|Anti-immigration||1.00 (0.21)***||0.50 (0.03)***|
|Authoritarianism||0.07 (0.03)*||0.07 (0.03)*|
|Traditionalism||0.01 (0.02)||0.01 (0.02)|
|Europe, globalization||-0.68 (0.04)***||-0.68 (0.04)***|
|Intercept||-3.43 (1.19)**||-1.35 (1.07)|
Logistic regressions, models with interactionsNotes: * p < 0.05; ** p < 0.01; *** p < 0.001; regression coefficients; standard errors in brackets. Vote in the first round of the 2017 presidential election; mainstream candidates: Hamon, Macron and Fillon. SCoRE survey data.
The socio-demographic and attitudinal variables in the full models (3a and 3b) are not presented in the tables in order to reduce the number of figures and enhance readability. However, they are included in the calculation of the estimates.
Hanspeter Kriesi, “The Populist Challenge,” West European Politics 37, no. 2 (2014): 361–78; Carlos De la Torre, The Promise and Perils of Populism: Global Perspectives (Lexington: The University Press of Kentucky, 2015); Ronald F. Inglehart and Pippa Norris, “Trump, Brexit, and the Rise of Populism: Economic Have-Nots and Cultural Backlash,” Faculty Research Working Paper Series, Harvard Kennedy School of Government, August 2016, RWP16–026; John B. Judis, The Populist Explosion: How the Great Recession Transformed American and European Politics, Columbia Global Reports, 2016; Giorgos Katsambekis, “The Populist Surge in Post-Democratic Times: Theoretical and Political Challenges,” The Political Quarterly 88, no. 2 (2017): 202–10.
This questions has been addressed in Yves Mény and Yves Surel, Par le peuple, pour le peuple: Le populisme et les démocraties (Paris: Fayard, 2000); Paul Taggart, Populism (Buckingham: Open University Press, 2000); Benjamin Moffitt and Simon Tormey, “Rethinking Populism: Politics, Mediatization and Political Style,” Political Studies 62, no. 2 (2014): 381–97; Jan-Werner Müller, What is Populism? (Philadelphia: University of Pennsylvania Press, 2016); Cas Mudde and Cristóbal Rovira Kaltwasser, Populism: A Very Short Introduction (New York: Oxford University Press, 2017).
Cas Mudde and Cristóbal Rovira Kaltwasser, Populism in Europe and the Americas: Threat or Corrective for Democracy? (Cambridge: Cambridge University Press, 2012).
In particular: Agnes Akkerman, Cas Mudde, and Andrej Zaslove, “How Populist are the People? Measuring Populist Attitudes in Voters,” Comparative Political Studies 47, no. 9 (2014): 1324–53; Eva Anduiza, Marc Guinjoan, and Guillem Rico, “Economic Crisis, Populist Attitudes, and the Birth of Podemos in Spain,” in Citizens and the Crisis, eds. Marco Giugni and Maria T. Grasso (Basingstoke: Palgrave Macmillan, 2018), 61–81; Mark Elchardus and Bram Spruyt, “Populism, Persistent Republicanism and Declinism: An Empirical Analysis of Populism as a Thin Ideology,” in Government and Opposition 51, no. 1 (2016): 111–33; Steven Van Hauwaert and Stijn Van Kessel, “Beyond Protest and Discontent: A Cross-National Analysis of the Effect of Populist Attitudes and Issue Positions on Populist Party Support,” European Journal of Political Research 57, no. 1 (2018): 68–92.
Catherine Colliot-Thélène and Florent Guénard, eds., Peuples et populisme (Paris: PUF, 2014); Mario E. Poblete, “How to Assess Populist Discourse through Three Current Approaches,” Journal of Political Ideologies 20, no. 2 (2015): 201–18.
Ghita Ionescu and Ernest Gellner, eds., Populism: Its Meanings and National Characteristics (New York: McMillan, 1969).
Jan Jagers and Stefaan Walgrave, “Populism as Political Communication Style: An Empirical Study of Political Parties’ Discourse in Belgium,” European Journal of Political Research 46, no. 3 (2007): 319–45; Benjamin Moffitt and Simon Tormey, “Rethinking Populism”; Paris Aslanidis, “Is Populism an Ideology? A Refutation and a New Perspective,” Political Studies 61, no. 1 (2016): 88–104; Andreu Casero-Ripollés, Marçal Sintes-Olivella, and Pere Franch, “The Populist Political Communication Style in Action: Podemos’s Issues and Functions on Twitter during the 2016 Spanish General Election,” American Behavioral Scientist 61, no. 9 (2017): 986–1001. On populism as a strategy, see Kurt Weyland, “Populism: A Political Strategic Approach,” in Cristóbal Rovira Kaltwasser, Paul Taggart, Paulina Ochoa Espejo, and Pierre Ostiguy, eds., The Oxford Handbook of Populism (Oxford: Oxford University Press, 2017), 48–72.
Kirk A. Hawkins, Ryan Carlin, Levente Littvay, and Cristóbal Rovira Kaltwasser, eds., The Ideational Approach to Populism: Concept, Theory, and Method (London: Routledge, 2018).
Cas Mudde, “The Populist Zeitgeist,” Government and Opposition 39, no. 4 (2004): 542–63.
Kirk Hawkins and Cristóbal Rovira Kaltwasser, “What the (Ideational) Study of Populism Can Teach Us, and What It Can’t,” Swiss Political Science Review 23, no. 4 (2017): 526–42.
Mudde and Kaltwasser, Populism in Europe and the Americas, 9–19.
Matthijs Rooduijn, “The Nucleus of Populism: In Search of the Lowest Common Denominator,” Government and Opposition 49, no. 4 (2014): 573–99.
Yannis Stavrakakis, Ioannis Andreadis, and Giorgos Katsambekis, “A New Populism Index at Work: Identifying Populist Candidates and Parties in the Contemporary Greek Context,” European Politics and Society 18, no. 4 (2017): 446–64, particularly 448–9.
Ben Stanley, “The Thin Ideology of Populism,” Journal of Political Ideologies 13, no. 1 (2008): 95–110, particularly 96.
Pierre Rosanvallon, “Penser le populisme,” in Colliot-Thélène and Guénard, eds., Peuples et populisme, 27–42.
Mény and Surel, Par le peuple, pour le peuple, 77.
Jean Leca, “Justice pour les renards! Comment le pluralisme peut nous aider à comprendre le populisme,” Critique 776–7 (2012): 85–95, particularly 91.
Margaret Canovan, “Trust the People! Populism and the Two Faces of Democracy,” Political Studies 47, no. 1 (1999): 2–16.
Mény and Surel, Par le peuple, pour le peuple, 39.
Mudde and Kaltwasser, Populism in Europe and the Americas, 18–19.
Mudde and Kaltwasser, Populism in Europe and the Americas, 6.
Recently Michael Freeden has emphasized the extremely “emaciated” nature of populism. Cf. Michael Freeden, “After the Brexit Referendum: Revisiting Populism as an Ideology,” Journal of Political Ideologies 22, no. 1 (2017): 1–11, particularly 3.
Mudde and Kaltwasser, Populism in Europe and the Americas; Matthijs Rooduijn and Tjitske Akkerman, “Flank Attacks: Populism and Left-Right Radicalism in Western Europe,” Party Politics 23, no. 3 (2017): 193–204; Matthijs Rooduijn, Brian Burgoon, Erika J. van Elsas, and Herman G. van de Werfhorst, “Radical Distinction: Support for Radical Left and Radical Right Parties in Europe,” European Union Politics 18, no. 4 (2017): 536–59; Gilles Ivaldi, Maria Elisabetta Lanzone, and Dwayne Woods, “Varieties of Populism Across a Left-Right Spectrum: The Case of the Front National, the Northern League, Podemos and Five Star Movement,” Swiss Political Science Review 23, no. 4 (2017): 354–76.
Herbert Kitschelt, The Transformation of European Social Democracy (Cambridge: Cambridge University Press, 1994); Liesbet Hooghe, Gary Marks, and Carole J. Wilson, “Does Left/Right Structure Party Positions on European Integration?,” Comparative Political Studies 35, no. 8 (2002): 956–89; for France, see Vincent Tiberj, “Two-axis Politics: Values, Votes and Sociological Cleavages in France (1988-2007),” Revue française de science politique 62, no. 1 (2012): 71–106.
Cas Mudde, Populist Radical Right Parties in Europe (Cambridge: Cambridge University Press, 2007).
Gilles Ivaldi, “Le Front national français dans l’espace des droites radicales européennes,” Pouvoirs 157 (2016): 115–126.
Stavrakakis and Katsambekis, “Left-Wing Populism in the European Periphery,” 129; Chantal Mouffe, “In Defence of Left-Wing Populism,” The Conversation, April 30, 2016, https://theconversation.com/in-defence-of-left-wing-populism-55869.
Gilles Ivaldi, “Parties and Voters in the Populist Market in France,” in Daniel Stockemer, ed., Populism around the World: A Comparative Perspective (Springer: Cham, 2019), https://www.worldcat.org/title/populism-around-the-world-a-comparative-perspective/oclc/1054092787.
Pascal Perrineau, ed., Les croisés de la société fermée: L’Europe des extrêmes droites (La Tour-d’Aigues: Éditions de l’Aube, 2001).
Hanspeter Kriesi, Edgar Grande, Romain Lachat, Martin Dolezal, Simon Bornschier, and Timotheos Frey, West European Politics in the Age of Globalization (Cambridge: Cambridge University Press, 2008).
Hanspeter Kriesi and Takkis S. Pappas, eds., European Populism in the Shadow of the Great Recession (Colchester: ECPR, 2015).
Enrique Hernández and Hanspeter Kriesi, “The Electoral Consequences of the Financial and Economic Crisis in Europe,” European Journal of Political Research 55 (2016): 203–24.
Catherine E. De Vries and Erica E. Edwards, “Taking Europe to Its Extremes: Extremist Parties and Public Euroscepticism,” Party Politics 15, no. 1 (2009): 5–28; Daphne Halikiopoulou, Kyriaki Nanou, and Sofia Vasilopoulou, “The Paradox of Nationalism: The Common Denominator of Radical Right and Radical Left Euroscepticism,” European Journal of Political Research 51, no. 4 (2012): 504–39.Online
Marianne Kneuer, “The Tandem of Populism and Euroscepticism: A Comparative Perspective in the Light of the European Crises,” Contemporary Social Science (2018); Céline Belot, Bruno Cautrès, and Sylvie Strudel, “Europe as a Polarising Issue: Its Effects on the Electoral Offer and Voting Preferences in the 2012 French Presidential Election,” Revue française de science politique 63, no. 6 (2013): 1081–112.
For a complete overview of this research, see Matthew Singer, Ioannis Andreadis, Kirk Hawkins, Ivan Llamazares, and Cristobal Rovira Kaltwasser, “The Conditional Effects of Populist Attitudes on Voter Choices in Four Democracies,” in Hawkins et al., The Ideational Approach to Populism.
The scale thus proposes a parsimonious measurement of a complex attitude. The empirical utility of this is corroborated by similar studies, some of which involve a larger number of items. See, for example, Anne Schulz, Philipp Müller, Christian Schemer, Dominique Stefanie Wirz, Martin Wettstein, and Werner Wirth, “Measuring Populist Attitudes on Three Dimensions,” International Journal of Public Opinion Research 30, no. 2 (2018): 316-26.
Akkerman et al., “How Populist Are the People?,” 1329; Stanley, “The Thin Ideology of Populism,” 102.
Guillem Rico and Eva Anduiza, “Economic Correlates of Populist Attitudes: An Analysis of Nine European Countries in the Aftermath of the Great Recession,” Acta Politica (2017).
In each of the two surveys, an exploratory principal component analysis confirmed the presence of a satisfactory solution with a single factor (% variance explained = 47%) and item loadings all above 0.5.
Mokken analysis is an exploratory method based on a generalization of the Guttman model, part of the family of item response theories (IRT), and is applicable to ordinal items. Unidimensionality tests are performed on a probabilistic basis depending on the level of “difficulty” of each item, including the possibility of errors (speeding, misunderstanding, random answers, etc.). Mokken models should also satisfy that the item response function is monotonically nondecreasing and the assumption of local independence of the items, i.e. that responses to the items are due to the individual’s position on the measured latent trait—items are no longer correlated when these correlations are controlled by the presence of the latent trait. See L. Andries van der Ark, “New Developments in Mokken Scale Analysis in R,” Journal of Statistical Software 48, no. 5 (2012): 1–27.
In the absence of comparable indicators, the rest of the article focuses on the SCoRE survey data.
For a recent literature review, see Gilles Ivaldi, “Electoral Basis of Populist Parties,” in Political Populism: A Handbook, edited by Reinhard C. Heinisch, Christina Holtz-Bacha, and Oscar Mazzoleni (Baden-Baden: Nomos, 2017), 157–68.
Martial Foucault, “La France politique des possédants et des non possédants”, in La démocratie de l’entre-soi, edited by Pascal Perrineau and Luc Rouban (Paris: Presses de Sciences Po, 2017), 43–56, particularly 54.
This is confirmed by Jérôme Fourquet, “Sur la gauche radicale: le vote Mélenchon,” Commentaire 159 (2017): 535–42.
Florent Gougou, “Les ouvriers et le vote Front national: les logiques d’un réalignement électoral,” in Les faux-semblants du Front national, edited by Sylvain Crépon, Alexandre Dézé, and Nonna Mayer (Paris: Presses de Sciences Po, 2015), 323–44; Pascal Perrineau, “Front national : un nouveau vote de classe?,” in Perrineau and Rouban, La démocratie de l’entre-soi, 133–46, particularly 144.
Benjamin Moffitt, “How to Perform Crisis: A Model for Understanding the Key Role of Crisis in Contemporary Populism,” Government and Opposition 50, no. 2 (2015) :189–217.
Hernández and Kriesi, “The Electoral Consequences”
Shaun Bowler, David Denemark, Todd Donovan, and Duncan Mcdonnell, “Right-Wing Populist Party Supporters: Dissatisfied but Not Direct Democrats,” European Journal of Political Research 56, no. 1 (2017): 70–91.
The relationship is complex: populism can occur within “pools” of voters without party identification, but many of these “non-partisan” voters remain quite largely immune to populism. See Carlos Meléndez and Cristóbal Rovira Kaltwasser, “Political Identities: The Missing Link in the Study of Populism,” Party Politics (2017).
These topics are measured from the following items. Subjective poverty: “We really can’t get by with household incomes” (scale of 1–4); political interest: “Would you say you are interested in politics?” (scale of 1–7); internal efficacy (sense of competence): “Politics is so complicated that you don’t really understand what’s going on” (scale of 1–7); external efficacy: “People like me don’t have much influence over what the government does” (scale of 1–7); party proximity: “Do you feel close to a political party” (yes/no).
We observed moderate correlations of 0.2 with subjective sense of poverty and -0.2 with external effectiveness. Collinearity diagnostics were conducted for models 2a and 2b to identify possible redundant variables that could distort estimates. However, examination of variance inflation factors (VIFs) does not indicate any serious multicollinearity, with all values below 2.
Average marginal effects (AMEs) are obtained by calculating the marginal effects for all possible values of the predictors, which are then averaged. They provide a more intuitive measure of the effect of a variable, which does not presuppose a fixed or mean value for the other parameters of the model. These effects were calculated with the “margins” package in R.
Note that the asymmetry of the distribution of the populism scale can be problematic. The relatively small number of individuals on the lowest scores of the scale can potentially distort the estimates. Sensitivity analysis is performed by removing individuals from the first and fourth quartiles of the populism scale, i.e. extreme values. On these truncated samples, the effect of populism remains positive and significant for the Mélenchon votes (coeff. = 0.87 p < 0.01; average marginal effect of 17%; N = 5,066) and Le Pen (coeff. = 1.39 p < 0.01; average marginal effect of 23%; N = 4,851).
The coefficients are positive and statistically significant: coefficients = 0.19 p < 0.01; N = 4,037 for the Mélenchon vote, contrasted with abstention; coefficients = 0.42 p <0.01; N = 3,857 for the Le Pen vote.
Céline Braconnier, Jean-Yves Dormagen, La Démocratie de l’abstention: Aux origines de la démobilisation électorale en milieu populaire, 2nd ed. (Paris: Gallimard, 2014). The SCoRE survey is no exception to the rule, particularly for those with the lowest education levels, who are underrepresented in the survey.
This is confirmed, particularly for right-wing populism, by the comparative study by Robert A. Huber and Saskia P. Ruth, “Mind the Gap! Populism, Participation and Representation in Europe?,” Swiss Political Science Review 23, no. 4 (2017): 462–84.
Richard Nadeau, Éric Bélanger, Michael Lewis-Beck, Bruno Cautrès, and Martial Foucault, Le vote des Français de Mitterrand à Sarkozy: 1988, 1995, 2002, 2007 (Paris: Presses de Sciences Po, 2012).
Rooduijn et al., “Radical distinction”; Vincent Tiberj, “Running to Stand Still: Le clivage gauche/droite en 2017,” Revue française de science politique 67, no. 6 (2017): 1089–112.
Ivaldi, “Electoral Basis.”
Hooghe et al., “Does Left/Right Structure Party Positions on European Integration?”
Perrineau, Les croisés de la société fermée. Bruno Cautrès shows that in France there is a “certain porosity” between opinions about European integration and globalization. See Bruno Cautrès, “La remise en cause de l’Europe,” in Perrineau and Rouban, La démocratie de l’entre-soi, 163–78, particularly 174.
Each of these dimensions forms a homogeneous attitude scale, with Cronbach’s alphas all above 0.6.
These effects persist for truncated samples (removing quartiles 1 and 4 from the populism scale): around 10% for Mélenchon voters (N = 3,629) and 7% for Le Pen (N = 3,522). While there are significant correlations between several attitudinal predictors, VIF values do not indicate any severe multicollinearity that could distort the estimates.
Mudde, and Kaltwasser, Populism in Europe and the Americas.
Van Hauwaert and Van Kessel, “Beyond Protest and Discontent,” 13.
Gérard Grunberg, “Le sombre avenir de la gauche,” in Le vote disruptif: Les élections présidentielle et législatives de 2017, ed., Pascal Perrineau (Paris: Presses de Sciences Po, 2017), 307–18.
It should be noted that populism is negatively correlated with individuals’ level of international openness (0.3).
Cautrès, “La remise en cause de l’Europe,” 173; Maurits Meijers, “Contagious Euroscepticism: The Impact of Eurosceptic Support on Mainstream Party Positions on European integration,” Party Politics 23, no. 4 (2017): 413–23.
Recall that we are still using the framework of multiple logistic regression models—that is, “neutralizing” the effect of all the other socio-demographic and ideological factors.
Nonna Mayer, “Les électeurs du Front national (2012–2015),” in La déconnexion électorale, edited by Florent Gougou and Vincent Tiberj (Paris: Fondation Jean-Jaurès, 2017), 69–76.
Éric Fassin, Populisme: Le grand ressentiment (Paris: Textuel, 2017), 83.
Van Hauwaert and Van Kessel, “Beyond Protest and Discontent,” 16.
Grunberg, “Le sombre avenir de la gauche,” 317; Emmanuelle Schön-Quinlivan, “The ‘Elephant in the Room’ No More: Europe as a Structuring Line of Political Cleavage in the 2017 Presidential Election,” French Politics 15, no. 3 (2017): 290–302.
Antonia María Ruiz Jiménez, Manuel Tomás González-Fernández, and Manuel Jiménez Sánchez, “Identifying with the Nation: Spain’s Left-Wing Citizens in an Age of Crisis,” South European Society and Politics 20, no. 4 (2015): 487–508; Mouffe, “In Defence of Left-Wing Populism.” See also the interview with Iñigo Errejón on this subject, http://lvsl.fr/macron-est-un-caudillo-neoliberal-entretien-avec-inigo-errejon.
Benjamin De Cleen and Yannis Stavrakakis, “Distinctions and Articulations: A Discourse Theoretical Framework for the Study of Populism and Nationalism,” Javnost: The Public 24, no. 4 (2017): 301–19.
J. Eric Oliver and Wendy M. Rahn, “Rise of the Trumpenvolk: Populism in the 2016 Election,” The ANNALS of the American Academy of Political and Social Science 667, no. 1 (2016): 89–206.
The author would like to thank Martial Foucault, Director of Cevipof, Sylvain Brouard, as well as the entire team in charge of the French Electoral Survey (ENEF) (https://www.enef.fr).