CAIRN-INT.INFO : International Edition
Volume
67
format_quote Cite share Share Switch to French

Full-text

1The electoral breakthrough of the populist radical right seen in Europe since the late 1980s was initially more pronounced among male voters than among female voters. [1] This phenomenon, termed the “Radical Right Gender Gap” (RRGG) by American researcher Terri Givens, was particularly apparent in France, where during Jean-Marie Le Pen’s time as head of the Front National (FN), gender was, along with education, the most strongly predictive factor for a vote in his favor. [2] This was no longer the case in 2012, when his daughter, who had just succeeded him as party leader, stood as a candidate in the presidential elections, and for the first time almost no gap was observed between the levels of support she received from men and women. But during all the midterm elections—European, regional, and at the département level—that have since taken place (all admittedly characterized by a high level of abstention), the gap has re-emerged. [3] This article aims to understand and explain these variations. It analyzes the impact of gender on the Marine Le Pen vote in the 2017 presidential election, in what was a particularly favorable context for her party as the Macron phenomenon swept aside the traditional parties, and support for radical parties grew dramatically at both ends of the political spectrum. Marine Le Pen won 21.3% of votes cast, progressing to the second round, and while her performance in the run-off fell far short of the 40% predicted by polling before her televised debate on 3 May, it was nevertheless higher than an FN candidate had ever previously achieved in a presidential election, with a total of 10.6 million votes, a share of 33.9%.

2Following an introduction that provides an overview of previous research in this area, our hypotheses, and the data used, the first section of this article investigates whether an RRGG was present in 2017, using a series of binary and multinomial logistic regressions controlling for the impact of gender with other sociodemographic and attitudinal variables typically used to explain the Le Pen/FN vote. It confirms the disappearance of the male-female differential over the two rounds of the presidential election. The second section then explores the generational dynamic of the vote in more detail, by looking at the intersection of gender with age, period, and birth cohort effects. It demonstrates that Marine Le Pen has succeeded in attracting and winning the loyalty of a new generation of women who reached voting age in 2012 and have only ever known the FN in the “detoxified” version embodied by its new leader. Using psychological indicators that have previously been seldom used in French electoral analyses—the “Big Five”—the third section explores the different psychological motivations of men and women for voting for Marine Le Pen. These five personality traits, which result from a complex interaction between heredity and environment, characterize the intellectual and affective make-up of an individual. [4] Combined with data on attitudes to authority and minorities, [5] they can be used to shed new light on our understanding of the FN vote, for example by showing that men with the lowest levels of emotional stability were more likely to vote for Marine Le Pen than other voters. However, these same personality traits had no impact on the female vote.

Explaining the Radical Right Gender Gap

An Overview

3Four broad explanations have been put forward to account for the relative disinclination of female voters to support the populist radical right. The first relates to gender differences in the labor market. Across the board, right-wing parties achieve their best results among working class voters: [6] a group of manual, primarily male laborers who are particularly exposed to unemployment and precarity and competition with immigrant workers. [7] In contrast, women generally perform non-manual labor, for example in the retail or service industries, or the public sector (teaching), and are thus thought to come into contact with immigrants less frequently and be less sensitive to xenophobic rhetoric. The second explanation highlights the role of religion, with Christian values supposedly acting as a shield against the inegalitarian and anti-universalist ideology of the far right. In France especially, the Catholic church has condemned FN policies on multiple occasions, and women—particularly older women—are in general more devout than men, and thus in principle more receptive to the teaching of the Church. The social spread of feminism has also been suggested as a third contributing factor, with young women in particular finding it hard to reconcile the emancipatory goals of feminism with right-wing values and a traditional view of the family, while conversely feminism may be seen as a threat by some men. [8] In contrast, despite the progress of ordinary feminism, the persistence of gender stereotypes is clear, with girls’ education promoting submission to norms, and discouraging aggressivity and assertiveness. [9] The image of extremism and violence associated with the far-right and its “outsider”, nonconformist side is therefore thought to dissuade female voters. [10]

4Recent work, however, has painted a more nuanced picture of the impact of these factors, showing that the extent of the RRGG greatly varies depending on the country and election in question. [11] During the 2012 French presidential election it disappeared, after controlling for the effects of other sociodemographic and political variables likely to influence it, and when gender and profession were combined the gap was in fact inverted, with employees in the primarily female retail sector more likely to vote for Le Pen than the primarily male category of manual laborers. [12] This would appear to suggest that the factors previously keeping women from supporting the radical right are being eroded.

5The boundaries between socio-professional categories are changing, with the tertiarization of manual labor and the proletarianization of non-manual jobs creating a hazily defined group of “unskilled workers” in which women are as numerous and as badly off as men. [13] Women working as supermarket cashiers, retail assistants, kindergarten staff, and cleaners embody an under-represented, under-recognized, and underpaid service proletariat with conditions of precarity that equal those of manual laborers, and encourage a retreat into the discourse of national identity. [14]

6The greater visibility of Islam in the public space, debates about headscarves and the burqa, and the growth of Islamic fundamentalism, have awakened identity-based discourse and a hardening of ethnocentric attitudes among French Catholics. After the Muhammad cartoon affair in particular, a previously non-existent correlation emerged between regularity of Catholic church attendance and levels of anti-immigrant prejudice. [15]

7Finally, the political offer has changed, with radical right parties evolving, trying to soften their extremist image, and communicating their discourse through euphemism. This has notably been the goal of the “detoxification” strategy led by Marine Le Pen since 2011. [16] Like other far-right leaders in Norway, Denmark, and the Netherlands, she has refocused her discourse on Islam, [17] presenting it as a threat to women’s rights and democracy.

New Hypotheses

8Based on this overview, for the 2017 election we put forward two sets of contrasting hypotheses: the first conjecturing a gradual disappearance of the gap between the level of FN support among men and women, and the second conjecturing the gap was maintained. These hypotheses stem from explanatory factors on both the demand and supply side of politics, and in particular from party positions on gender equality, a topic that has been little studied in the literature to date. [18]

9The hypothesis of a disappearing RRGG is supported by the increasingly precarious professional situation of women following the 2008 crisis, particularly in the service sector, which has made the working conditions of primarily female non-manual workers similar to those of manual laborers, a predominantly male group. [19] It is further supported by the erosion of the shield of Catholic values, in a context of competing identity discourses, and with the Catholic church apparently divided on how to deal with the FN. On the eve of the 2017 presidential run-off vote, the French Conference of Catholic Bishops merely called upon voters to “recall the fundamental principles”, unlike in 2002 when it had clearly urged voters to stand in the way of the FN candidate.

10Other explanatory factors favoring the disappearance of the RRGG, this time on the side of the political offer, include of course the “Marine Le Pen” effect. Le Pen, who has been the party leader for the last six years, presents herself as a “modern” woman who, in contrast to her father’s deliberately sexist and masculine rhetoric, claims to understand women who have abortions, has welcomed gay members into the party, emphasizes her identity as a divorced and remarried mother, [20] and has even gone as far as describing herself in her biography as a “quasi-feminist”. [21] In 2017, she made a particular effort to present herself as a candidate who understood the needs of female voters, a strategy described by Catherine Achin and Sandrine Lévêque as “gendered normalization”. [22] Her manifesto focused on equal pay, the fight against working precarity, and most importantly the fight against Islamism, “which rolls back women’s fundamental freedoms”, [23] and she even produced a four-page leaflet, written in an intimate and personal tone, that was dedicated especially to women. [24] Such tactics are particularly significant because, as shown by Eelco Harteveld and his colleagues, men and women do not respond to the same issues in party programs, particularly when it comes to the populist radical right. [25] We also hypothesize that the strategy of winning over women would be particularly influential among new generations of female voters, [26] who reached voting age when Marine Le Pen was already FN leader and her normalization strategy was making the headlines. [27] Unfamiliar with her father, they first experienced French political life at a time when Marine Le Pen was already an omnipresent figure. These facilitating factors were particularly likely to have had an impact in 2017, when, as in 2012, the elections were for the presidential office and thus particularly personality-focused and motivating—elections in which the Marine Le Pen effect could be felt to its full extent.

11The second set of hypotheses, on the other hand, is based on the factors (on the side of political demand) likely to keep female support for the FN and its leader lower. Chief among these is the fact that women’s education encourages them to internalize social norms, particularly in relation to racism and ethnocentrism, to a greater extent than men. The experiments conducted by Elisabeth Ivarsflaten and Eelco Harteveld show that women are more motivated than men to control anti-immigrant prejudice, and to prioritize the democratic norm of tolerance. [28]

12To explore this hypothesis in greater depth, this article draws upon work in political psychology. After a period of neglect, the study of personality and its impact on political behavior is making a powerful comeback. [29] It is primarily based on personality traits, which describe “what people are like” [30]—in contrast to values, which reflect what individuals consider to be important—and are relatively stable throughout a person’s life. [31] They affect perceptions, values, behavior, and the way in which individuals interact in society. [32] Five major factors, identified by Ernest Tupes and Raymond Christal in 1961 and popularized by Lewis R. Goldberg as the so-called “Big Five”, are considered to summarize them reasonably well. [33] They are “agreeableness”, “extraversion”, “emotional stability”, [34] “conscientiousness”, and “openness to experience”. [35]

13Agreeableness is synonymous with caring, the ability to make friends, and capacity for empathy. Extraversion indicates warmth, sociability, and assertiveness. Emotional stability refers to an individual’s ability to control negative emotions such as anxiety, depression, and anger. Conscientiousness reflects the level of “socially prescribed impulse control”, [36] describing self-controlled personalities, with a sense of organization, rigor, and method. Openness to experience indicates openness to diversity, and an interest in new experiences and culture. [37] Each of these personality traits is measured by individual scores on a scale (see appendix). And, notably, they are unevenly distributed between the sexes, with women appearing to be more extrovert, more sociable, and more conscientious than men, while men appear to be more emotionally stable.

14These personality traits may influence support for the FN candidate since, among other things, they influence attitudes toward minorities and relationships to authority: attitudes that play a key role in support for the radical right. [38] Agreeableness and openness to experience might therefore act as a barrier to the Marine Le Pen vote by shielding against ethnocentrism, while conscientiousness might favor authoritarianism. As women tend to be more open in relation to these indicators than men, they might therefore be less likely to vote for the FN. However, because they are also more subject to social control, they might also be more favorable to authoritarianism. In other words, they combine traits that push in opposite directions in relation to the likelihood of voting for Le Pen.

15Finally, these long-term factors might have been compounded by a short-term factor: Marine Le Pen’s very poor performance during the televised debate held between the two rounds of the election, on 3 May, in which she appeared particularly aggressive and rude toward Emmanuel Macron, contradicting the calm and softer image that her “detoxification” strategy had sought to create. This brutal change in register is likely to have had a greater impact on women, who are more sensitive than men to social norms and more reticent about resorting to violence. We hypothesize that this might favor the re-emergence of an RRGG in the run-off.

16To verify these hypotheses, we have used data from the French Electoral Study (FES 2017), which was carried out after the presidential election run-off. [39] Demonstrating the gradual normalization of the Front National, declared votes in favor of Le Pen in the first round were close to actual results, underestimated only by -0.7 points (though 5 points in the run-off) and we therefore have substantial numbers for analysis (291 Le Pen voters in the first round, and 336 in the run-off).

The Radical Right Gender Gap Disappears Again

17The survey confirms the disappearance of the gender gap during the 2017 presidential election. The presidential elections remobilized female voters, who had been more likely than male voters to abstain from voting during previous midterm elections. [40] In 2017, however, more women voted than men, particularly in the run-off (table 1). This was confirmed by the data regardless of the indicator used, whether it was based on responses to a direct question about participation in each round, formulated so as to facilitate the admission of abstention, [41] or whether participation was estimated indirectly, with “no votes” considered as the sum of declared abstentions, blank or spoiled ballot papers, and refusals to respond. Women did not only abstain from voting less frequently, but were also less likely to submit a blank or spoiled ballot paper (3.8% of female voters compared to 5% of male voters in the first round, and 15.1% compared to 19% in the run-off).

Table 1

Voter Participation in the 2017 Presidential Elections (in Percentages)*

Declared participationTotal votes cast*
1st roundRun-off1st roundRun-off
Men83.378.775.360.3
Women86.582.079.066.4
Gap+3.2+3.3+3.7+6.1

Voter Participation in the 2017 Presidential Elections (in Percentages)*

* Excluding declared abstentions, blank and spoiled ballot papers, and refusal to respond.
Source: FES 2017. Unweighted data.

18In total, support from female voters therefore played a particularly strong role in this election, representing 53.4% of votes cast in the first round, and 54.5% in the run-off (compared to 52.4% of registered voters at the end of 2016, according to Insee).

19Compared to men (table 2), in the first round women were distinguished by lower levels of support for the far left and the traditional right (-2.8 points for the Jean-Luc Mélenchon vote and -2.3 for the François Fillon vote), and by greater support for Benoît Hamon and Marine Le Pen (+3.2 and +2.6 points respectively). In the run-off, there was almost no difference. Not only did the RRGG disappear, but it actually tended to invert, although the gaps were small (figure 1).

Table 2

Voter Choices in the Two Rounds of the 2017 Presidential Elections (as a Percentage of Votes Cast)

1st round (N = 1,414)Run-off (N = 1,162)
Far leftMélenchonHamonMacronFillonLe PenOthersMacronLe Pen
Men1.721.14.624.421.120.2765.834.2
Women1.718.37.823.318.822.87.365.634.4
Gap0– 2.8+ 3.2– 1.1– 2.3+ 2.6+ 0.3– 0.2+ 0.2

Voter Choices in the Two Rounds of the 2017 Presidential Elections (as a Percentage of Votes Cast)

Source: FES 2017. Data weighted based on the respective results of the 1st and 2nd rounds.
Figure 1

Change in the Le Pen Vote by Women and Men in Presidential Elections (as a Percentage of Votes Cast)*

Figure 1

Change in the Le Pen Vote by Women and Men in Presidential Elections (as a Percentage of Votes Cast)*

* Based on the first round, except for 2017. Data unavailable for votes by gender in the 2002 run-off. Source: Cevipof French post-electoral surveys 1988 and 1995; French electoral panel 2002 and 2007 (wave 1), FES 2012 and 2017 (CEE); data adjusted based on actual results.

20These averages may however be misleading, masking composition effects that reflect the impact of other gender characteristics that may influence voting behavior. A series of logistic regressions, introducing sociodemographic control variables (age, education, individual profession, religious observance) (table 3), then political attitudes (position on the left-right spectrum, scores on Euroskepticism, ethnocentrism, economic and cultural liberalism scales [42]) (table 4) allowed us to investigate whether gender had a specific impact on the Marine Le Pen vote in the first round and run-off of the presidential election, [43] all other things remaining equal. For the first round, we conducted a multinomial regression analysis to compare reasons for voting for Marine Le Pen with reasons for voting for the main candidates, Jean-Luc Mélenchon, Emmanuel Macron, and François Fillon, compared to all other choices (vote for the seven remaining minor candidates or non-vote), first with the sociodemographic variables only, then with the attitudinal variables. For the run-off, a binary logistic regression compared the Le Pen vote with the Macron vote. [44]

Table 3

Multinomial Logistic Regression on Votes for the Primary Candidates in the First Round of the 2017 Presidential Election Based on Sociodemographic Variables

Mélenchon B(se)Macron B(se)Fillon B(se)Le Pen B(se)
Gender *
Men0.158 (0.15)-0.131(0.15)-0.022(0.18)-0.201(0.16)
Women–
Age
18-240.158(0.28)– 1.124(0.31)***– 1.484(0.46)**0.783(0.31)*
25-340.252(25)– 1.219(0.27)***– 1.611(0.37)***06576(0.28)*
35-490.059(0.23)– 0.743(0.21)***– 0.832(0.25)***0.824(0.25)**
50-640.173(0.22)– 0.061(0.19)– 0.787(0.23)***0.610(0.23)***
65 and over
Education level
Primary, no qualifications– 0.780(0.29)**– 1.231(0.27)***– 1.003(0.33)**1.091(0.29)***
BEPC (9th grade diploma)– 0.602(0.29)*– 1.141(0.31)***– 0.424(0.35)0.451(0.32)
CAP (Vocational diploma)– 0.711(0.20)***– 1.015(0.231)***– 0.805(0.28)**0.536(0.24)*
Vocational baccalaureate– 0.407(0.24)– 0.208(0.24)– 0.072(0.32)0.730(0.27)**
General baccalaureate– 0.288(0.25)– 0.496(0.26)– 0.015(0.32)– 0.792(0.42)
Higher education
Profession
Inactive0.245(0.28)0.059(0.26)– 0.748(0. 37)*0.975(0.39)*
Self-employed– 0.047(0.35)– 0.343(0.32)0.522(0.31)1.229(0.42)**
Middle-ranking professions0.341(0.25)– 0.005(0.23)– 0.798(0.30)**1.094(0.38)**
Non-manual workers0.385(0.25)– 0.428(0.24)– 0.781(0.29)**0.964(0.37)*
Manual workers0.219(0.28)– 0.292(0.27)– 1.037(0.36)**0.887(0.38)*
Senior managers
Religion
Catholic (regular attendance)– 0.939(0.44)*0.220(0.33)2.422(0.35)***– 0.060(0.38)
Catholic (occasional attendance)– .685(0.30)*0.325(0.25)1.785(0.32)***0.492(0.25)
Non-practicing– 0.372(0.15)*0.348(0.16)*1.309(0.26)***0.184(0.17)
Other religion– 0.087(0.22)– 0.411(0.29)0.495(0.42)– 1.198(0.39)**
No religion
Constant– 0.438(0.27)0.474(0.24)– 0.763(0.33)*– 2.719(0.40)***
N1830183018301830
Log-likelihood3112.1123112.1123112.1123112.112
X2 (df)528.594 (76)***528.594(76)***528.594(76)***528.594(76)***
R20.2630.2630.2630.263

Multinomial Logistic Regression on Votes for the Primary Candidates in the First Round of the 2017 Presidential Election Based on Sociodemographic Variables

Source: FES 2017. Unweighted data. * p < 0.05; ** p < 0.010; *** p < 0.001. The reference category is a vote for the seven other candidates and non-cast votes.
Table 4

Multinomial Logistic Regression on Votes for the Main Candidates in the First Round of the 2017 Presidential Election, Including Attitudinal Variables

Mélenchon B(se)Macron B(se)Fillon B (se)Le Pen B(se)
Gender
Men0.209(0.16)– 0.170(0.16)– 0.266.21– 0.302(0.19)
Women
Age
18-240.116(0.31)– 1.221(0.33)***– 1.1400(0.50)*1.280(0.37)**
25-340.383(26)– 1.153(0.28)***– 1.582(0.41)***0.876(0.32)**
35-490.102(0.24)– 0.816(0.23)***– 0.896(0.30)**1.157(0.29)***
50-640.215(0.23)– 0.027(0.20)– 0.581(0.26)*0.646(0.26)*
65 and over
Education level
Primary, no qualifications– 0.532(0.33)– 0.577(0.30)– 0.838(0.39)*0.330(0.35)
BEPC (9th grade diploma)– 0.503(0.32)– 0.713(0.32)*– 0.470(0.40)– 0.163(0.38)
CAP (Vocational diploma)– 0.564(0.22)*– 0.521(0.23)*– 0.621(0.32)– 0.020(0.28)
Vocational baccalaureate– 0.318(0.26)0.124(0.26)0.133(0.36)0.461(0.34)
General baccalaureate– 0.272(0.27)– 0.290(0.27)– 0.006(0.36)– 1.091(0.47)*
Higher education
Profession
Inactive0.426(0.30)0.400(0.28)– 0.575(0.41)0.486(0.44)
Self-employed0.338(0.38)– 0.384(0.34)0.346(0.35)0.962(0.46)*
Middle-ranking professions0.546(0.27)0.215(0.24)– 0.533(0.33)0.782(0.42)
Non-manual workers0.613(0.27)*– 0.172(0.25)– 0.462(0.33)0.657(0.41)
Manual workers0.315(0.30)– 0.063(0.28)– 0.490(0.40)0.662(0.43)
Senior managers
Religion
Catholic (regular attendance)– 0.758(0.48)0.282(0.34)1.847(0.39)***– 0.223(0.43)
Catholic (occasional attendance)– 0.255(0.32)0.330(0.26)1.312(0.36)***0.402(0.29)
Non-practicing– 0.019(0.17)0.359(0.17)*0.954(0.28)**0.074(0.20)
Other religion0.107(0.26)– 0.435(0.31)0.80(0.47)– 1.193(0.46)*
No religion
Attitudes
Position on L-R spectrum– 0.328(0.04)***– 0.031(0.03)0.616(0.06)***0.363(0.4)***
Anti-EU sentiment– 0.019(0.03)– 0.257(0.03)***– 0.177(0.04)***.0161(0.3)***
Ethnocentrism– 0.079(0.02)***– 0.030(0.02)0.030(0.03)0.219(0.2)***
Cultural liberalism– 0.042(0.03)0.041(0.03)– 0.054(0.04)– 0.035(0.3)
Economic liberalism– 0.194(0.04)***0.184(0.04)***0.377(0.05)***0.018(0.4)
Constant2.408(0.84)**0.363(0.82)– 5.931(1.10)***– 8.611(1.01)***
N1830183018301830
Log-likelihood4180.5614180.5614180.5614180.561
X2 (df)1445.740(96)***1445.740(96)***1445.740(96)***1445.740(96)***
R20.5730.5730.5730.573

Multinomial Logistic Regression on Votes for the Main Candidates in the First Round of the 2017 Presidential Election, Including Attitudinal Variables

Source: FES 2017, unweighted data. * p < 0.05; ** p < 0.010; *** p < 0.001. The reference category is a vote for the seven other candidates and non-cast votes.

21The result was clear. Regardless of the control variables incorporated into the model, in both the first and second rounds, gender had no statistically significant electoral impact in 2017. In the first round, as in 2012, the male/female differential disappeared, and contrary to our expectations it did not return in the run-off, despite Marine Le Pen’s aggressive performance in the televised debate held between the two rounds. Her success is explained by variables other than gender, in particular education, age, political orientation, and views on immigrants and the European Union. Compared to those who supported her opponents, Marine Le Pen’s 2017 voters appear to have been highly typical. She primarily attracted right-wing, overwhelmingly ethnocentrist, and anti-European voters. She won over not necessarily young, but active voters (aged under 65), but was rejected by more educated voters (those with a general baccalaureate or higher), those with a higher social status (senior management, liberal, and intellectual professions), and those from a minority religion, in particular Muslims. Jean-Luc Mélenchon attracted voters who were most left-wing on the left-right spectrum, the least ethnocentrist, and the most hostile to economic liberalism, and had strong support from non-manual workers. Support for Emmanuel Macron was associated with strong pro-European sentiment and a high level of economic liberalism. He did not attract young voters or those with a low level of education (BEPC, CAP, or vocational diploma). François Fillon attracted very right-wing voters who had liberal economic views and were strongly pro-European, and, like Macron, was unpopular with young and uneducated voters.

22To round out our observations, we must consider possible interactions between gender and other explanatory factors. In 2012, the variables predicting a vote for Marine Le Pen were exactly the same, regardless of gender. But separate binary logistic regressions for women and men, based on the same variables used in this article, [45] showed that these variables had a consistently lower impact on the female vote, particularly attitudinal variables. [46] Thus, all other things being equal, the predicted likelihood of a vote for Marine Le Pen was greatest among individuals furthest to the right on the left-right spectrum, but reached 79% among men, compared to 51% among women. This RRGG only disappeared among the most Euroskeptic individuals, and inverted to a small extent (-2) in the presence of a strong declared sympathy for the FN candidate. In 2012, in other words, Marine Le Pen had already managed to overcome the reticence of part of the female electorate who sympathized with the ideas of the FN, but hesitated to take the final step.

23In 2017, the same variables still predicted the Marine Le Pen vote in the first round, among both women and men. But the gender effect was reversed: at the far-right, anti-European, and strongly ethnocentric end of the spectrum, women consistently voted for Marine Le Pen not less, but more frequently, than men. And while sympathy for Marine Le Pen still played a stronger role than all other variables, gaps between the genders on this indicator were smaller than those generated by the other attitudes considered. The picture is therefore more complex than it first appears.

The Le Pen Vote, Gender, and Generational Renewal

24The regression models presented in the first section of this article have two limitations. They do not sufficiently account for possible interactions between age and gender, and they do not enable an understanding of the “age” variable in all its complexity. Age combines three dimensions that may each have different effects: the moment in the life cycle, the current period—here the specific context of the election in question—and the birth cohort, i.e. the fact of having been born at the same time and having been marked by the same social, political, and economic events. These three temporal factors (age, period, and cohort) cannot be handled separately in an analysis that does not use longitudinal data. [47] In this second section, which focuses in particular on the effects of period and cohort and their interactions with gender, we therefore seek to deconstruct the effects of age using longitudinal data from French presidential election surveys from 1988 to 2017. [48]To test our hypotheses, we created “age-by-period” tables that enabled us to produce descriptive statistics distinguishing between the effects of age, period, and cohort. [49] Tables 5 and 6 show the proportions of women and men per age category voting for Jean-Marie Le Pen and then for his daughter in the last six presidential elections in France. The columns show the Le Pen vote by age at each election, and the vote of each age category can be tracked over time by reading along the rows. Finally, reading the table diagonally makes it possible to identify possible cohort effects, seeing for example how people who were aged 18-26 during the 1988 elections voted seven years later, in the 1995 election when they were aged 25-33, then seven years later when they were aged 32-40, and so on. However, this analysis has two limitations: first, the 2000 reform of the length of the presidential mandate changed the gap between each election from seven to five years, meaning that some cohorts overlap age categories in the two tables from 2007 onward; and second, since the gaps between elections have changed, the final age group is very broad, covering three decades of births for each period. [50] Individuals who were 48 years old in 1988, belonging to the cohort born in 1940, can therefore be found in the “67 and older” age category for the 2007, 2012, and 2017 elections. Despite these flaws, some trends can be identified from these two tables.

Table 5

Proportion of Women Voting Le Pen by Cohort, Age Group, and Election (as a Percentage of Votes Cast)

Period
Age198819952002200720122017Average
18-268.715.912.9920.331.916.5
27-368.41611.19.622.325.715.5
37-469.913.217.110.31520.714.4
47-569.8820.612.414.727.515.5
57-668.411.2166.619.719.513.6
67 and older7.47.410.86.913.416.210.4
Total for women8.812.314.89.117.522.8

Proportion of Women Voting Le Pen by Cohort, Age Group, and Election (as a Percentage of Votes Cast)

Source: French electoral surveys 1988, 1995, 2002, 2007 (Cevipof); FES 2012 and 2017 (CEE). Data adjusted based on the actual presidential election results.
Table 6

Proportion of Men Voting Le Pen by Cohort, Age Category, and Election (as a Percentage of Votes Cast)

Period
Age198819952002200720122017Average
18-2612.121.410.39.12624.917.3
27-3613.721.722.210.420.528.719.5
37-4612.919.118.514.719.124.218.1
47-5615.621.522.410.225.419.219.1
57-6611.814.724.315.116.714.516.3
67 and older13.112.314.114.310.416.113.4
Total for men13.118.61912.319.220.1

Proportion of Men Voting Le Pen by Cohort, Age Category, and Election (as a Percentage of Votes Cast)

Source: French electoral surveys 1988, 1995, 2002, 2007 (Cevipof); FES 2012 and 2017 (CEE). Data adjusted based on actual presidential elections results.

25First, in regard to age as a moment in the life cycle, tables 5 and 6 show overall that these effects are not linear over time—in either men or women. Contrary to popular opinion, the FN vote is not consistently a “young” vote. And while an age effect and a “vote surge” for the FN was apparent among the youngest voters in 1995, among both men and women, this was not the case in 2002, where Jean-Marie Le Pen was most popular with women aged 47 to 56 years and with men aged 47 to 66 years. This was an “elderly” vote, in an election dominated by the theme of insecurity. [51] The same was true in 2007, in a low ebb for the FN, which found itself no longer able to attract the support of new generations. [52] The FN vote was primarily maintained in the intermediate age groups, regardless of gender. Our regression models for the two rounds in 2017 (first section) concluded that there was a strong age effect, ceteris paribus. A descriptive longitudinal analysis of “age-period-cohort” effects qualifies this result (tables 5 and 6) by demonstrating the emergence of major fluctuations linked to elections and their particular related issues.

26Second, tables 5 and 6 demonstrate a clear period effect. Among both women and men, the Le Pen vote grew between 1988 and 2002, dropped dramatically in 2007 due to the “Sarkozy effect”, and rose again between 2012 and 2017. [53] Aside from the dip in 2007, we can see an overall rise in the Le Pen vote from one election to the next, across all age groups and regardless of gender, but this period effect was not exactly the same for women and men. Among male voters, the Le Pen vote in the 18-26 age category doubled between 1988 and 2017, but quadrupled among women of the same age group during the same period. In 1988, the FN won 8.7% of votes among the youngest female voters, while 29 years later, it won 31.9% of their support. This “gender-period” effect can also be seen in the older age groups: between 1988 and 2017, the proportion of the Le Pen vote among women aged 47 to 56 years tripled, while among men in the same age category it did not even double. And while FN support among the oldest male voters only increased by 1 to 3 points between 1988 and 2017, it doubled during the same period among the oldest female voters. Our results clearly show a period effect here, which is conditioned by gender. In other words, over time the FN has made up for its deficit among female voters across all age groups. The context of each presidential election has certainly had an impact, more marked among women than men. And as the differences between men and women essentially blur, or even invert, from 2012 onward (table 7), the decisive variable underpinning this gender-period effect would appear to be the replacement of Jean-Marie Le Pen as party leader by his daughter. Our results are in line with the hypothesis of a “Marine Le Pen effect”. An inversion of the gender gap began to form in 2012, but was truly confirmed in 2017 (table 7), where it could be seen across almost all age groups and almost all cohorts.

Table 7

Male/Female Differential for the FN Vote by Election and Age Group (as a Percentage of Votes Cast)

Period
Age198819952002200720122017
18-263.45.5– 2.60.15.7– 7
27-365.35.711.10.8– 1.83
37-4635.91.44.44.13.5
47-565.813.51.8– 2.210.7– 8.3
57-663.43.58.98.5– 3– 5
67 and older5.74.93.37.4– 3– 0.1
Total difference4.36.34.23.21.7-2.7

Male/Female Differential for the FN Vote by Election and Age Group (as a Percentage of Votes Cast)

Interpretation: the table shows the differences between the proportion of men and women voting Le Pen by age category. A negative difference indicates that more women than men voted for Le Pen.
Source: French electoral surveys 1988, 1995, 2002, 2007 (Cevipof); FES 2012 and 2017 (CEE). Data adjusted based on actual presidential election results.

27However, two particular cohorts stand out: the two most recent, who voted for the first time in 2012 or in 2017 (tables 5-7). Support from young French women, which since 1988 had always remained below the threshold of 16%, grew substantially in 2012 and truly broke through in 2017. Nearly 32% of women aged 18 to 26 voted for Marine Le Pen in the last presidential election: a much higher proportion than their male counterparts. The gender gap for this age category inverted and rose to 7 percentage points (table 7). In this young cohort, new to electoral politics and born in the late 1980s or the early 1990s, which has only known the FN as led by Marine Le Pen, the Le Pen vote has become a very largely female vote: an inverted Radical Right Gender Gap. In other words, in the first round of the 2002 presidential election, women represented 43.5% of Le Pen voters, in 2012 they represented half, and in 2017 they accounted for 56% of the Le Pen vote. [54]

Gender, Personality Traits, and the Le Pen Vote among Men and Women

28In contrast, the factors likely to make women still reluctant to support a radical right party like the FN include personality traits. Women, as we have seen, appear to be more inclined to self-control, and are therefore more respectful of social convention and less drawn to extreme and nonconformist behavior. This may reflect a personality trait and/or the different ways in which boys and girls are socialized. At the personality level, this can be verified using the Big Five personality instrument described above. The FES 2017 survey included a simplified version, the Ten Item Personality Inventory (TIPI) (see appendix). Descriptive statistics show that on average, women score slightly higher on agreeableness and extraversion, while men have higher levels of emotional stability. To what extent did these traits influence the Marine Le Pen vote?

29To answer this question, we ran a series of logistic regressions that assessed the explanatory potential of the Big Five [55] for the Marine Le Pen vote in the first round of the 2017 presidential election (table 8). They first tested the impact of the Big Five alone (model 1), then added sociodemographic variables (model 2), and ethnocentrist and authoritarian attitudes (model 3). We then carried out separate regressions for men (model 4) and women (model 5). Finally, we tested the effect of the interaction between gender and personality traits across the entire electorate (model 6).

Table 8

Logistic Regression on the Marine Le Pen Vote in the First Round of the 2017 Presidential Election Including Personality Traits

Model 1 B(es)Model 2 B(es)Model 3 B(es)Model 4 (M) B (se)Model 5 (W) B(se)Model 6 B (se)
Big Five
Extraversion– 0.028(0.06)– 0.024(0.06)– 0.016(0.07)– 0.053(0.12)0.021(0.10)0.021(0.10)
Agreeableness0.086(0.07)– 0.014(0.07)– 0.070(0.09)– 0.510(0.13)– 0.085(0.12)– 0.073(0.12)
Conscientiousness0.091(0.06)0.124(0.07)0.052(0.08)0.150(0.13)– 0.045(0.12)– 0.012(0.11)
Emotional stability– 0.134(0.05)*– 0.078(0.05)– 0.080(0.06)– 0.247(0.10)*0.043(0.09)0.043(0.08)
Openness to experience– 0.025(0.06)– 0.050(0.06)0.158(0.07)*0.269(0.12)*0.079(0.10)0.081(0.10)
Big Five*Gender [56]
Extraversion*Gender– 0.069(0.15)
Agreeableness*Gender– 0.005(0.18)
Conscientiousness*Gender0.169(0.17)
Emotional stability*Gender– 0.299(0.13)*
Openness to experience*Gender0.168(0.15)
Attitudes
Ethnocentrism0.312(0.02)***0.337(0.04)***0.302(0.03)***0.312(0.02)***
Authoritarianism0.540(0.08)***0.546(0.12)***0.605(0.12)***0.560(0.08)***
Gender
Men– 0.122(0.15)–– 0.183(1.37)
Women–0.083(0.17)–
Age
18-241.555(0.29)***1.915(0.35)***1.876(0.49)***2.002(0.53)***1.915(0.35)***
25-341.422(0.27)***1.725(0.32)***1.654(0.48)**1.803(0.45)***1.749(0.32)***
35-491.317(0.24)***1.743(0.28)***1.698(0.43)***1.922(0.40)***1.760(0.28)***
50-640.658(0.21)**0.862(0.25)**0.766(0.37)*0.969(0.35)**0.861(0.25)**
65 and over–––––
Level of education
Primary, no qualifications2.419(0.26)***1.277(0.31)***1.102(0.45)1.373(0.46)**1.269(0.32)***
BEPC (9th grade diploma)1.415(0.30)***0.500(0.36)– 0.006(0.64)0.838(0.47)0.549(0.36)
CAP (Vocational diploma)1.676(0.21)***0.747(0.25)**0.497(0.37)0.993(0.37)**0.783(0.20)**
Vocational baccalaureate1.280(0.24)***0.655(0.28)*– 0.171(0.42)1.395(0.41)**0.660(0.28)*
General baccalaureate– 0.325(0.40)– 0.773(0.46)0.034(0.69)– 1.109(0.65)– 0.758(0.46)
Higher education––––
Religion
Catholic (regular attendance)– 0.360(0.37)– 0.636(0.43)– 1.290(0.82)– 0.160(0.53)– 0.601(0.43)
Catholic (occasional attendance)0.415(0.24)0.006(0.29)**– 1.105(0.52)*0.603(0.37)0.001(0.29)
Non-practicing0.110(0.16)– 0.171(0.20)– 0.340(0.30)– 0.044(0.28)– 0.173(0.20)
Other religion– 1.071(0.40)**– 0.927(0.46)*– 0.766(0.58)– 1.775(1.06)– 0.973(0.46)*
No religion––––
Constant– 1.464(0.54)**– 3.90(0.64)***– 8.465(0.85)***– 8.598(1.33)***– 8.590(1.18)***– 8.394(1.04)***
N1414141414146507641414
Log-likelihood– 713.862– 627.036– 455.437– 200.238– 240.337– 451.936
X2 (df)9.85(5)183.51(19)***526.70(21)***230.16 (20)***324.24(20)***533.65(26)***
R20.0060.1270.36660.3650.4020.371

Logistic Regression on the Marine Le Pen Vote in the First Round of the 2017 Presidential Election Including Personality Traits

Source: FES 2017. Unweighted data. With *p < 0.05; **p < 0.010; ***p < 0.001.

30Contrary to our initial hypotheses, neither conscientiousness, which favors conformism and social conservatism, nor agreeableness, which discourages individuals from controversial behavior, such as voting for a radical candidate, appear to have had an impact on the Marine Le Pen vote. Emotional instability did have a direct impact on the Le Pen vote, even after controlling for sociodemographic and attitudinal factors, but only among men (models 4 and 5), while a higher degree of openness to experience increased the likelihood of voting for her (model 3), but again only among men (models 4 and 5).

31The effect of openness to experience, which initially appears counterintuitive, may in fact make sense, as open-minded personalities are not necessarily more tolerant to minorities but instead more inclined to embrace novelty of any kind. As such, with her anti-establishment stance and radical ideas, Marine Le Pen strikes a chord with a certain group of voters who see her as an “unconventional” choice reflecting their own personality. For women this appeal might be counteracted by their stronger internalization of social norms. However, the relationship between openness to experience and voting for Marine Le Pen is in fact statistically significant only in the regression carried out for male voters (model 4), and disappears when we test the interaction between openness and gender for the electorate as a whole (model 6).

32The impact of emotional stability, on the other hand, appears to be more robust. Logistic regressions for male voters (model 4) and female voters (model 5) show that emotional instability increased the likelihood of voting for Marine Le Pen, but only among men, all other things being equal. This effect was confirmed even after introducing the interaction between gender and personality traits (model 6). Nevertheless, the level of ethnocentrism and authoritarianism was by no means one of the most determining factors for this vote. [57]

33Finally, the impact of personality traits on the Marine Le Pen vote varied depending on gender. Although on average women have a higher degree of openness to experience, this trait did not influence their vote. Emotional instability had no impact on the likelihood of women voting for Marine Le Pen, and played a role only among men. [58] Although the RRGG appears to have disappeared in 2017, it remains that the Le Pen vote among men and women cannot be entirely explained by the same psychological traits.

34* * *

35These results are preliminary, and must be further explored. First, by transcending the some what sterile debate about whether the RRGG exists, and exploring how gender influences other electoral decisions on both the left and the right. Alongside the RRGG there is also, particularly in the intermediate generations, a “modern” gender gap with higher numbers of women voting for left-wing, particularly socialist parties, and a “traditional” gender gap with higher numbers of conservative voters. Each of these gender gaps has its own genera tional dynamic, which cannot be outlined here for reasons of space. [59] From an intersectionalist perspective, we also need to conduct a systematic comparison of cohort effects on electoral behavior with those of social class, religion, education, and origin. And the economic, cul tural, and moral values underlying these generational divides must also be considered, as in Vincent Tiberj’s Les citoyens qui viennent and in Florent Gougou’s analysis of the working class vote during the Fifth Republic. [60] Finally, we need to investigate whether the erosion in the RRGG is still seen in France in the period between presidential elections, during less candidate-centered and less mobilizing elections, and whether it is observed in other countries.

36Our current data do however enable us to state that the Radical Right Gender Gap is being eroded in France. During the last two presidential elections, a statistically significant difference could no longer be observed between overall levels of support for the FN candidate from male and female voters. Further still, in the young cohorts that have reached voting age since 2012, and have only known the FN under Marine Le Pen’s leadership, the malefemale differential was inverted. In 2017, the Marine Le Pen vote reached 32% in the first round among female voters under 26 years of age, i.e. 7 points higher than among their male counterparts. In this age group, the level of support for the radical right embodied by the FN quadrupled in comparison to 1988, while among men it only doubled. The FN leader’s strategy for winning over the female electorate has paid off, allowing her not only to make up on but even invert the RRGG among the youngest women. Even Marine Le Pen’s poor performance in the debate with Emmanuel Macron between the two rounds, in which she appeared both aggressive and incompetent, did not impede this catch up process, but merely slowed it down, with a slightly higher proportion of female voters still evident among the youngest cohort in the run-off (+0.3%, compared to 7% in the first round). Finally, the fact that personality traits, despite strongly shaping the way in which we think and act, had no direct impact on electoral behavior reminds us once again of the importance of factors relating to the political offer: the candidates standing, the issues featured, and the specific context of the election. The gendered dimension of these factors, such as the impact of the style and rhetoric of female leaders, and the role and positioning of gender issues in the political program of radical right parties, has as yet been little studied in the literature, but would appear to be vital in view of our results. Marine Le Pen has twice succeeded in overcoming the reticence of the female electorate to vote for her, particularly among working class women (manual and non-manual workers), by presenting a different image to that of her father and addressing them directly. It remains to be seen—if she is re-elected leader during the FN rebuilding conference set to be held in March 2018— whether she will continue to make such progress. [61]

Appendix

Attitude Scales

37Left-right spectrum from 0 “very left-wing” to 10 “very right-wing”, recoded 1-11. “Don’t know” answers were given the scale’s mean value of 6 (average score of respondents on the scale).

38Sympathy for Marine Le Pen from 0 “don’t like her at all” to 10 “like her a lot”, recoded 1-11. “Don’t know” answers were given the scale’s mean value of 3 (average score of respondents on the scale).

39Cultural liberalism (4-16). Sum of the responses to four questions recoded in growing permissiveness, average of sum of responses: “homosexuality is an acceptable way of expressing one’s sexuality”, “homosexual couples should be able to adopt children”, “a woman’s most important role is to have children and raise them”, “women should have the right to choose to have an abortion” (from “completely agree” to “completely disagree”).

40Ethnocentrism (4-19). Sum of the responses to four questions recoded in growing intolerance, average of sum of responses: “there are too many immigrants in France”, “in general French culture is threatened by immigrants”, “immigrants cause the crime rate to go up”, “many immigrants come to France solely to benefit from social security” (from “completely agree” to “completely disagree” for the first item with an additional option for the three others, “neither agree nor disagree”).

41Economic liberalism (4-13). Sum of the responses to four questions recoded in growing liberalism, average of sum of responses: “in order to face economic difficulties do you think the state should trust companies and give them more freedom, or instead control and regulate them more closely?”, “do you think that higher priority should be given in the coming years to the competitiveness of the French economy, or to improving the conditions of workers?”, “the government should take measures to reduce income inequality”, “to bring about social justice, we must take from the rich to give to the poor” (from “completely agree” to “completely disagree”).

42Anti-European sentiment (4-15). Sum of the responses to four questions recorded in growing rejection of the European Union, average of sum of responses: “do you think that it is a good thing, a bad thing, or neither good nor bad for France to be part of the European Union?”, “all things considered, do you think that France has benefited or not benefited from belonging to the European Union? Yes, it has benefited/No, it has not benefited”, “decision-making powers should be strengthened in Europe even if this means reducing French independence”, “European integration currently prevents democracy from working properly in France” (from “completely agree” to “mostly disagree”, and the option “neither agree nor disagree”).

43The Big Five

44The Ten-Item Personality Inventory (TIPI) is based on the following question: “Here are a number of personality traits that may or may not apply to you. Please write a number next to each statement to indicate the extent to which you agree or disagree with that statement. You should rate the extent to which the pair of traits applies to you, even if one characteristic applies more strongly than the other. I see myself as: 1/ extraverted, enthusiastic; 2/ critical, quarrelsome; 3/ dependable, self-disciplined; 4/ anxious, easily upset; 5/ open to new experiences, complex; 6/ reserved, quiet; 7/ sympathetic, warm; 8/ disorganized, careless; 9/ calm, emotionally stable; 10/ conventional, uncreative”. There are seven types of response to indicate the extent to which the statements apply: “1/ disagree strongly”; “2/ disagree moderately”; “3/ disagree a little”; “4/ neither agree nor disagree”; “5/ agree a little”; “6/agree moderately”; “7/agree strongly”.

45Each personality trait is measured by two pairs of items, one measuring the trait and the other its opposite, negatively correlated. For example, an extraverted individual will have a high score for item pair 1 (“extraverted, enthusiastic”) and a low score for pair 6 (“reserved, quiet”). Respondents’ individual scores for each of the five traits are therefore measured as a sum of the score given to the first item (from 1 to 7) and the inverse score (1=7; 2=6, etc). of the second pair.

Table 9

Calculating Scores for the Big Five

Big FivePair of items usedMethod of calculating the Big Five
Extraversion1/ Extraverted, enthusiastic
6/ Reserved, quiet (recoded to the inverse scale)		Extraversion score = score for item 1 + score for inverse item 6 / 2
Agreeableness2/ Critical, quarrelsome (recoded to the inverse scale)
7/ Sympathetic, warm		Agreeableness score = score for inverse item 2 + score for item 7 / 2
Emotional stability4/ Anxious, easily upset (recoded to the inverse scale)
9/ Calm, emotionally stable		Emotional stability score = score for inverse item 4 + score for item 9 / 2
Openness to experience5/ Open to new experiences, complex
10/ Conventional, uncreative (recoded to the inverse scale)		Openness to experience score = score for item 5 + score for inverse item 10 / 2
Conscientiousness3/ Dependable, self-disciplined
8/ Disorganized, careless (recoded to the inverse scale)		Conscientiousness score = score for item 3 + score for inverse item 8 / 2

Calculating Scores for the Big Five

46The coefficient alpha for all item pairs making up each of the Big Five is weak (below 0.50). However, this coefficient tends to underestimate the reliability of ordinal scales, and the very nature of the TIPI battery used in our survey is a simplified battery with only two items to measure each personality trait. [62]

Table 10

Average Score for the Big Five by Gender

AgreeablenessExtraversionConscientiousnessEmotional stabilityOpenness to experience
Women (W)5,363,945,524,705,00
Men (M)5,153,695,244,294,90
W-M difference+ 0,21+ 0,25+0,28– 0,41– 0,10

Average Score for the Big Five by Gender

Notes

  • [1]
    See in particular Hans-Georg Betz’s seminal work Radical Right-Wing Populism in Western Europe, New York, Saint Martin’s Press, 1994. For a recent overview of the political impact of gender, see the special issue “Gender and Populist Radical-Right Politics”, Patterns of Prejudice, 49(1-2), 2015.
  • [2]
    Terri Givens, “The Radical Right Gender Gap”, Comparative Political Studies, 37(1), 2004, 30-54.
  • [3]
    See in particular Nonna Mayer, “The closing of the Radical Right Gender Gap in France?”, French Politics, 13(4), 2015, 391-414, and “Les électeurs du Front national (2012-2015)” in Florent Gougou and Vincent Tiberj (eds), La déconnexion électorale: Un état des lieux de la démocratie française, Paris, Fondation Jean-Jaurès, 2017, pp. 69-76; Réjane Sénac and Maxime Parodi, “‘Gender Gap à la française’: recomposition ou dépassement?”, Revue Française de Science Politique, 63(2), 2013, 225-48.
  • [4]
    For an introduction to the use of personality traits in analyzing political behavior, see Alan S. Gerber, Gregory A. Huber, David Doherty, and Conor M. Dowling, “The Big Five personality traits in the political arena”, The Annual Review of Political Science, 14, 2011, 265-87.
  • [5]
    Two attitudes at the heart of the radical right-wing vote, as shown by Cas Mudde in Populist Radical Right Parties in Europe, Cambridge, Cambridge University Press, 2007, 23.
  • [6]
    See in particular Daniel Oesch, “Explaining voters’ support for right-wing populist parties in Western Europe: evidence from Austria, Belgium, France, Norway and Switzerland”, International Political Science Review, 29(3), 2008, 348-73; Simon Bornschier and Hanspeter Kriesi, “The populist right, the working class, and the changing face of class politics” in Jens Rydgren (ed.), Class Politics and the Radical Right, Abingdon, Routledge, 2012, pp. 10-29; Dennis Spies, “Explaining working-class support for extreme right parties: a party competition approach”, Acta Politica, 48, 2013, 296-325.
  • [7]
    On the complex relations between French workers and immigrants in factories, see the investigation by Stéphane Beaud and Michel Pialoux, Retour sur la condition ouvrière: Enquête aux usines Peugeot de Sochaux-Montbéliard, Paris, La Découverte, 2012.
  • [8]
    Pascal Perrineau, Le Symptôme Le Pen: Radiographie des électeurs du Front national, Paris, Fayard, 1997, 105-7.
  • [9]
    On the issue of the gendered education of girls (and boys), see Barrie Thorne, Gender Play: Girls and Boys in School, New Brunswick, Rutgers University Press, 1993; Anne Dafflon Novelle (ed.), Filles-garçons: Socialisation différenciée?, Grenoble, Presses Universitaires de Grenoble, 2006; Muriel Darmon, La socialisation, Paris, Armand Colin, 2006.
  • [10]
    On the greater conformity of women with anti-racist norms, see Eelco Harteveld and Elisabeth Ivarsflaten, “Why women avoid the radical right: internalized norms and party reputations”, British Journal of Political Science, 48(2), 2018, 369-84. On their lower inclination to vote for socially stigmatized parties, see the experiments carried out by Eelco Harteveld with Wouter Van Der Brug, Stefan Dahlberg, and Andrej Kokkonen, “The gender gap in voting: extremity, ambiguity and social cues”, ECPR conference, Bordeaux, 5-7 September 2013.
  • [11]
    Tim Immerzeel, Hilde Coffé, and Tanja Van der Lippe, “Explaining the gender gap in radical right voting: a cross-national investigation in 12 Western-European countries”, Comparative European Politics, 13(2), 2015, 263-86.
  • [12]
    Ibid.
  • [13]
    Thomas Amossé and Olivier Chardon, “Les travailleurs non qualifiés: une nouvelle classe sociale?”, Économie et Statistique, 393-4, 2006, 203-29. See also Martin Thibault, Ouvriers malgré tout: Enquête sur les ateliers de maintenance des trains de la Régie autonome des transports parisiens, Paris, Raisons d’agir, 2013.
  • [14]
    Amossé and Chardon, “Les travailleurs”, 222-4.
  • [15]
    Nonna Mayer and Guy Michelat, “Les transformations du rapport à l’autre: le rôle des identités politiques et religieuses” in Commission nationale consultative des droits de l’homme: La Lutte contre le racisme, l’antisémitisme et la xénophobie, Paris, La Documentation française, 2007, 122-38; Guy Michelat and Claude Dargent, “Système catholique symbolique et comportements électoraux”, Revue française de science politique, 65(1), 2015, 27-60. However, since the 2013 election of Pope Francis, who embodies a more open kind of Catholicism, this correlation has disappeared.
  • [16]
    Alexandre Dézé, “‘La dédiabolisation’: Une nouvelle stratégie?” in Sylvain Crépon, Alexandre Dézé, and Nonna Mayer (eds), Les faux-semblants du Front national: Sociologie d’un parti politique, Paris, Presses de Sciences Po, 2015, pp. 27-50.
  • [17]
    On the feminization of the immigration issue for these parties, see Sarah de Lange and Liza Mügge, “Gender and right-wing populism in the Low Countries: ideological variations across parties and times”, Patterns of Prejudice, 49(1-2), 2015, 61-80. On the FN, see Sylvain Crépon, “La politique des murs au Front national” in Crépon et al. (eds) Les faux-semblants, pp. 185-205.
  • [18]
    Susi Meret, “Charismatic female leadership and gender: Pia KjÒrsgaard and the Danish People’s Party”, Patterns of Prejudice, 49(1-2), 2015, 81-102. The author shows that party positions on gender issues remain underexplored in studies of the impact of gender on the populist radical right vote, as do the impact of the rhetoric, style, charisma, and discourse of female populist leaders.
  • [19]
    According to the 2016 Insee Job Survey, the population aged 15 years and older includes 76% women in non-manual jobs, a proportion rising to 86% in the service sector, and 80% men in manual labor, rising to 87% among skilled manual workers (“Population selon le sexe et la catégorie socioprofessionnelle en 2017”, https://www.insee.fr/fr/statistiques/2381478, last accessed 13 August 2018).
  • [20]
    Mariette Sineau, “Vote/comportements politiques” in Catherine Achin and Laure Bereni (eds), Dictionnaire: Genre et science politique, Paris, Presses de Sciences Po, 2013, pp. 526-7.
  • [21]
    Marine Le Pen, “contre flots, Paris, Grancher, 2006, 188. Her feminism is however entirely relative, as shown by Catherine Achin and Sandrine Lévêque, “‘Jupiter is Back’: gender in the 2017 French presidential campaign”, French Politics, 15(3), 2017, 279-89.
  • [22]
    See Achin and Lévêque, “‘Jupiter is Back’”, 284.
  • [23]
    See her “144 presidential commitments”, including commitment no. 9, https://www.rassemblementnational.fr/le-projet-de-marine-le-pen/, last accessed 13 August 2018.
  • [24]
    Sylvain Chazot, “Karine Le Marchand, la féminité et des oublis… ce tract un tout petit peu surprenant de Marine Le Pen”, https://lelab.europe1.fr/karine-le-marchand-la-feminite-et-des-oublis-ce-tract-un-tout-petit-peusurprenant-de-marine-le-pen-2969387, last accessed 13 August 2018.
  • [25]
    Eelco Harteveld, Wouter Van Der Brug, Stefan Dahlberg, and Andrej Kokkonen, “The gender gap in populist radical-right voting: examining the demand side in Western and Eastern Europe”, Patterns of Prejudice, 49(1-2), 2015, 103-34.
  • [26]
    On the impact of gender and generation on electoral behavior in France, see Mariette Sineau, “Effets de genre, effets de génération? Le vote hommes/femmes à l’élection présidentielle 2007”, Revue française de science politique, 57(3), 2007, 353-69.
  • [27]
    See Alexandre Dézé, “La construction médiatique de la ‘nouveauté’ FN” in Crépon et al. (eds), Les faux-semblants, 455-504.
  • [28]
    Harteveld and Ivarsflaten, “Why women avoid the radical right”. On the indicator of “implicit motivation to control prejudice” (IMCP), see Scott Blinder, Robert Ford and Elisabeth Ivarsflaten, “The better angels of our nature: how the antiprejudice norm affects policy and party preferences in Great Britain and Germany”, American Journal of Political Science, 57(4), 2013, 841-57.
  • [29]
    Dans Shang E. Ha, Seokho Kim, and Se Hee Jo, “Personality traits and political participation: evidence from South Korea”, Political Psychology, 34(4), 2013, 511-32, particularly 512.
  • [30]
    Markus Freitag and Kathrin Ackermann, “Direct democracy and institutional trust: relationships and differences across personality traits”, Political Psychology, 37(5), 2016, 707-23, particularly 710.
  • [31]
    Gerber et al., “The Big Five”.
  • [32]
    Ha et al., “Personality traits”.
  • [33]
    Gerber et al., “The Big Five”; Lewis R. Goldberg, “The development of markers for the big-five factor structure”, Psychological Assessment, 4(1), 1992, 26-42.
  • [34]
    This is also sometimes referred to as “neuroticism”, its opposite. See Gerber et al., “The Big Five”.
  • [35]
    On the issue of the different facets of each personality trait and their translation into French, see: Odile Plaisant, Julie Guertault, Robert Courtois, Christian Réveillère, Gerald A. Mendelsohn, and Oliver P. John, “Histoire des ‘Big Five’: OCEAN des cinq grands facteurs de la personnalité. Introduction du Big Five Inventory français ou BFI-Fr”, Annales médico-psychologiques, 168(7), 2010, 481-6.
  • [36]
    Gerber et al., “The Big Five”, 267.
  • [37]
    Harald Schoen and Markus Steinbrecher, “Beyond total effects: exploring the interplay of personality and attitudes in affecting turnout in the 2009 German federal election”, Political Psychology, 34(4), 2013, 533-52, here 535.
  • [38]
    For more detail on use of the Big Five in analyzing political behavior, see Gerber et al., “The Big Five”. See also Matthias Fatke, “Personality traits and political ideology: a first global assessment”, Political Psychology, 38(5), 2017, 881-99.
  • [39]
    The study was led by Nicolas Sauger at the Sciences Po Center for European Studies and Comparative Politics, and conducted by Kantar TNS using face-to-face interviews, from a national sample built on quotas (gender, age, education level, profession of the reference person, and regions) of 1,830 individuals representative of the registered electorate residing in mainland France, carried out in the two weeks following the run-off of the 2017 presidential election. Tables showing the votes are adjusted based on the actual results for each round and the sociodemographic structure. Regressions were however carried out using unweighted data. The data were made available in July. The results presented here are therefore the very earliest available. See Florent Gougou and Nicolas Sauger, “The 2017 French Election Study (FES 2017): a post-electoral cross-sectional survey”, French Politics, 15(3), 2017, 360-70.
  • [40]
    Mayer, “Les électeurs”, 75.
  • [41]
    “You did not go and vote, you thought about going to vote, but in the end you did not go, you usually vote, but this time you did not go/you voted” (only one individual refused to answer this question out of the 1,830 individuals in the sample).
  • [42]
    See the appendix for information on these scales.
  • [43]
    For a comparison of previous elections with the same variables based on binary logistic regressions, see Mayer, “The closing of the Radical Right Gender Gap”, and Mayer, “Les électeurs”.
  • [44]
    For reasons of space this is not included in the article, but is available upon request.
  • [45]
    In all our regressions gender was used as a proxy for the gender variable. For a more general discussion of the handling of gender in statistical surveys in the social sciences, see Laurel Westbrook and Aliya Saperstein, “New categories are not enough: rethinking the measurement of sex and gender in social surveys”, Gender & Society, 29(4), 2015, 534-60.
  • [46]
    Mayer, “The closing of the Radical Right Gender Gap”, 405-6. For reasons of space, sympathy for Marine Le Pen is not included in the multinomial logistic regression model for 2017 (table 4). The results are available upon request. The inclusion of this variable increases the explanatory power of the model (the Nagelkerke R2 increases to 0.66), strongly increasing the likelihood of voting for the FN candidate, all other things being equal, and conversely reducing the likelihood of voting for Jean-Luc Mélenchon, Emmanuel Macron, and in particular François Fillon.
  • [47]
    Even where longitudinal data is available, no regression model can handle these three variables together (age, period, and cohort), as they are strictly collinear. It is therefore formally impossible to model their respective effects without imposing restrictions on the model. On the issue of so-called “APC” analyses, see Anja Neundorf and Richard Niemi, “Beyond political socialization: new approaches to age, period, cohort analysis”, Electoral Studies, 33, 2014, 1-6.
  • [48]
    French electoral surveys 1988, 1995, 2002, 2007 (Cevipof); FES 2012 and 2017 (CEE).
  • [49]
    Yang Yang and Kenneth C. Land, Age-Period-Cohort Analysis: New Models, Methods, and Empirical Applications, Boca Raton, CRC Press, 2013.
  • [50]
    This recoding is necessary due to the low number of observations in this category.
  • [51]
    Various media stories, including for example the attack on Paul Voise, marked the background to this election. See Nonna Mayer, Ces Français qui votent Le Pen, Paris, Flammarion, 2002, 351-5.
  • [52]
    Nonna Mayer, “Comment Nicolas Sarkozy a rétréci l’électorat Le Pen”, Revue française de science politique, 57(3), 2007, 429-45.
  • [53]
    Ibid.
  • [54]
    Proportions calculated based on adjusted data.
  • [55]
    For the significance threshold and methods of calculating the Big Five from the Ten Items Personality Inventory (TIPI) used in the FES 2017, see the appendix.
  • [56]
    The reference category is women.
  • [57]
    We checked to see whether there were any multicollinearity issues, i.e. strong correlations between personality traits and ethnocentric and authoritarian attitudes, but this was not the case. All the “tolerance” coefficients for the attitudinal variables and the Big Five of the model were greater than 0.70, i.e. well above the threshold of 0.10 that would indicate multicollinearity. We also explored possible interactions between personality traits and attitudes, which were inconclusive and not shown here due to lack of space. They require further testing using more in-depth statistical methods.
  • [58]
    Additional tests show that openness to experience and conscientiousness also influence levels of ethnocentrism and authoritarianism, even after controlling for age, gender, education, and religiousness. In other words, these personality traits would appear to have an indirect impact on attitudes through an effect of intermediation in line with our initial hypothesis. The results of these tests are available from the authors upon request.
  • [59]
    Results of the “age-by-period” tables for the other right-wing and left-wing political parties are available upon request.
  • [60]
    Vincent Tiberj, Les citoyens qui viennent: Comment le renouvellement générationnel transforme la politique en France, Paris, Presses Universitaires de France, 2017, in particular 236-41 on the values driving the FN vote in new cohorts; Florent Gougou, “Comprendre les mutations du vote des ouvriers: vote de classe, transformation des clivages et changement électoral en France et en Allemagne depuis 1945”, PhD. diss., 2012, Sciences Po Paris; Florent Gougou and Nonna Mayer, “The class basis of extreme right voting in France: generational replacement and the rise of new cultural issues (1984-2007)” in Jens Rydgren (ed.), Class Politics and the Radical Right, Abingdon, Routledge, 2012, pp. 156-72.
  • [61]
    This article owes a great deal to its anonymous reviewers, to whom we offer our thanks. We would also like to thank Florent Gougou for providing us with access to his cumulative data files of the French Electoral Surveys.
  • [62]
    On the problematic issue of using the coefficient alpha for ordinal scales, see Anne M. Gadermann, Martin Guhn, and Bruno D. Zumbo, “Estimating Ordinal Reliability for Likert-Type and Ordinal Item Response Data: A Conceptual, Empirical, and Practical Guide”, Practical Assessment Research & Evaluation, 17(3), 2012, 1-13.

Outline

  1. Explaining the Radical Right Gender Gap
    1. An Overview
    2. New Hypotheses
  3. The Radical Right Gender Gap Disappears Again
  4. The Le Pen Vote, Gender, and Generational Renewal
  5. Gender, Personality Traits, and the Le Pen Vote among Men and Women

Authors

Abdelkarim Amengay
Abdelkarim Amengay is a PhD candidate supervised at the Sciences Po Centre for European Studies and Comparative Politics and the School of Political Studies at the University of Ottawa. His dissertation studies the impact of media content on votes for the French FN. In 2015 he co-authored (with Daniel Stockemer) “The Voters of the FN under Jean-Marie Le Pen and Marine Le Pen: Continuity or Change?”, French Politics, 13(4), 2015, 370-90. More broadly, his research interests include electoral behaviour, radical right-wing parties, and political psychology (CEE, Sciences Po Paris, 27 rue Saint-Guillaume, 75337 Paris cedex 07.
abdelkarim.amengay@sciencespo.fr
Anja Durovic
Anja Durovic is a PhD candidate at the Sciences Po Centre for European Studies and Comparative Politics. Her dissertation is focused on the comparative analysis of gender inequalities in political participation in Europe, and the way in which generational change and contrasting institutional settings may influence the evolution of these inequalities. She recently published: “A Longitudinal Analysis of Gendered Patterns in Political Action in France: A Generational Story?”, French Politics, 15(4), 2017, 418-42. She is also part of the French team involved in the international research project: Gendered Electoral Financing in Democratic and Democratizing States (CEE, Sciences Po Paris, 27 rue Saint-Guillaume, 75337 Paris cedex 07.
anja.durovic@sciencespo.fr
Nonna Mayer
Nonna Mayer is emerita research professor at the French National Center for Scientific Research (CNRS), affiliated to the Sciences Po Centre for European Studies and Comparative Politics. She recently published (with Mickäel Durand) “Genre, sexualité et vote” in Yves Déloye and Nonna Mayer (eds), Analyses électorales, Brussels, Bruylant, 2017, pp. 265-318; and co-edited (with Alexandre Dézé and Sylvain Crépon) Les faux-semblants du Front national: Sociologie d’un parti politique, Paris, Presses de Sciences Po, 2015. Her research focuses on political attitudes and behaviors, right-wing extremism, and new forms of racism and anti-semitism (CEE, Sciences Po Paris, 27 rue Saint-Guillaume, 75337 Paris cedex 07.
nonna.mayer@sciencespo.fr
This is the latest publication of the author on cairn.
Latest publication on cairn or another partner portal
Abstract French
Latest publication on cairn or another partner portal
Abstract French
Uploaded on Cairn-int.info on 25/07/2019
Cite
Distribution électronique Cairn.info pour Presses de Sciences Po © Presses de Sciences Po. Tous droits réservés pour tous pays. Il est interdit, sauf accord préalable et écrit de l’éditeur, de reproduire (notamment par photocopie) partiellement ou totalement le présent article, de le stocker dans une banque de données ou de le communiquer au public sous quelque forme et de quelque manière que ce soit.
keyboard_arrow_up
Chargement
Loading... Please wait