A new party system in the making? The 2017 French presidential election

Transcription

1 Fr Polit (2017) 15: DOI /s ORIGINAL ARTICLE A new party system in the making? The 2017 French presidential election Florent Gougou 1 Simon Persico 1 Published online: 21 August 2017 Ó Macmillan Publishers Ltd 2017 Abstract The 2017 presidential election represents a turning point in French electoral politics. It was marked by the poor performance of the main governing parties candidates and the victory of political newcomer Emmanuel Macron. More surprisingly, four candidates with clearly distinct policy lines were neck and neck in the first round. This article sheds light on this outcome and assesses its consequences for the French party system. We sketch alternative scenarios regarding the format and content of the emerging party system. Using geometrical analyses on data from the French Election Study 2017 (Gougou and Sauger 2017), we show that the current political space is structured by two main conflict dimensions: the first and dominating dimension sets an anti-immigration/authoritarian pole against a proimmigration/libertarian pole; the second pits an ecologist/interventionist pole against a productivist/neoliberal pole. Keywords French elections Presidential election Party system Tripolarity Cleavages Introduction The election of Emmanuel Macron, a 39-year-old liberal candidate who was not supported by any of the traditional parties, was unexpected until only a few months before the election. His victory marks a major break with previous electoral patterns under the Fifth Republic and for two main reasons. First, almost all former & Florent Gougou florent.gougou@iepg.fr Simon Persico simon.persico@iepg.fr 1 Sciences Po Grenoble, PACTE, Grenoble, France

2 304 F. Gougou, S. Persico Presidents had a long political career in one or other of the two traditional governing parties in France, the Socialist Party and the Republicans (formerly UMP, RPR and UDR). Second, although Macron won the second round with a 30-point margin over Marine Le Pen, the first-round result was extremely close, with four candidates winning between 19.6 and 24% of the vote. This tight result was all the more significant because it represented four extremely distinct policy alternatives. Marine Le Pen defended the standard National Front manifesto, speaking out against immigration and economic globalization. Emmanuel Macron took a moderately liberal policy stance, in terms of cultural values (pro-immigration, pro-gay rights, etc.) as well as economic development (pro-austerity, pro-business). The Republican candidate, François Fillon, championed neoliberal reforms and social conservatism (on gay rights and French identity). Finally, the radical left contender, Jean-Luc Mélenchon, managed to merge proposals from the traditional left (regarding wealth redistribution and government intervention in the economy) and the green movement (against nuclear energy and in favor of alternative economic development). On April 23, 2017, the French political space seemed to be divided into four different blocks. This paper assesses the consequences of the 2017 electoral earthquake (Martin 2017) for the French party system. Despite Emmanuel Macron s clear victory, the closeness of the first-round result led many pundits and scholars to suggest that the new French party system was a four-party system ( quadripartition ). Our article examines this hypothesis along with three alternative ones: the standard tripolarity hypothesis, where there is one left-wing block, one moderate right-wing block and one extreme-right block; a new tripolarity, with one eco-socialist block on the left, one pro-globalization neoliberal block at the center and one anti-immigration conservative block on the right; and the rebirth of the quadrille bipolaire (Duverger 1985) organized around two poles each composed of two parties. This article addresses the classical literature on party systems, by focusing on the connection between voters and candidates rather than other aspects of party systems, e.g., the number of effective parties or the inter-party relationships. Our goal is to assess how voters fit into the new political space in terms of issues and values. This means evaluating the connection between citizens preferences and values and their choice on 23 April. In other words, we will assess what issues differentiate one voter from one another and which value set leads to one candidate rather than another. In order to map the 2017 French political space, we rely on geometric data analysis (GDA) and use data from the French Election Study 2017 (Gougou and Sauger 2017). The 2017 presidential election result Table 1 provides the results of the election. Voter turnout in the first round was relatively high (77.8%) even though it marked a slight 1.7 percentage point decrease compared to the previous presidential election. This was the second-lowest turnout for the first round of a French presidential election. From this perspective, the figure was in line with the long-term decline that has been visible in every type of election in France over 30 years (Gougou and Labouret 2011, p. 241). That said, voter turnout was unquestionably strong from a comparative perspective: the

3 A new party system in the making? The 2017 French 305 Table 1 Results of the 2017 French presidential election. Source: Constitutional Council. Results are percentages of the electorate for voters (turnout) and valid votes, and the percentage of the valid votes for each candidate Metropolitan France French Republic First round Second round First round Second round Electorate 44,443,212 44,430,041 47,582,183 47,568,693 Voters 35,563, ,883, ,003, ,467, Valid votes 34,689, ,941, ,054, ,381, Nathalie Arthaud LO 219, , Philippe Poutou NPA 376, , Jean-Luc Mélenchon FI-PCF 6,803, ,059, Benoît Hamon PS-EELV 2,189, ,291, Emmanuel Macron EM 8,267, ,677, ,656, ,743, Jacques Cheminade DIV 62, , Jean Lassalle DIV 427, , François Fillon LR-UDI 6,899, ,212, François Asselineau UPR 313, , Nicolas Dupont-Aignan DLF 1,665, ,695, Marine Le Pen FN 7,465, ,264, ,678, ,638,

4 306 F. Gougou, S. Persico figure was 68.5% in the first round of the 2016 Austrian presidential election, 66.5% in the 2016 Spanish general election, 81.9% in the 2017 Dutch general election, 68.8% in the 2017 British general election. The French presidential election is still able to mobilize the vast majority of the French electorate. Voter turnout was not critical to understand the new patterns of electoral competition. Turnout was not equally distributed among social groups, though. According to the French Election Study, abstention was 13 percentage point higher among young voters (18 24 age group) than elderly people (over 65 s); 12 percentage point higher among workers than among the managerial and professional occupations; 10 percentage point higher among people with only a basic education than those with a university degree. Turnout rates were consistently higher in upper-class constituencies than in lower-class constituencies. Voter turnout reached 87% in the 5th arrondissement of Paris, a central district where the Sorbonne University is situated, but only 60.8% in Clichy-sous-Bois, a poor suburb of Paris in Seine-Saint- Denis and 69% in Saint-Dizier, a former industrial town in Haute-Marne (Eastern France). Such patterns were said to disadvantage Marine Le Pen and Jean-Luc Mélenchon compared to François Fillon and Emmanuel Macron. Turnout in the second round fell to 74.8%. Historically, turnout tends to increase between the two rounds. Two key features explain why turnout actually decreased: the low competitiveness of the election and the absence of a left-wing competitor. This had already been the case in the 1969 presidential election when the runoff featured two right-wing candidates, Alain Poher and future President Georges Pompidou. As a matter of fact, a large fraction of left-wing voters declined to choose between Emmanuel Macron and Marine Le Pen, leading to a record number of blank invalid ballots as well (8.6% of the registered electorate). The first round was very close with four candidates literally neck and neck. This was the first presidential election where there had been such a narrow gap between the second and the fourth candidates. That said, previous presidential elections had already featured a four-way fight, with an 8.3 percentage point difference between the frontrunner and the fourth candidate in 1995 and a 9.3 percentage point difference in What was different this time was the failure of the traditional big governing parties, the Socialist Party and the Republicans. For the first time since 1981, neither mainstream party won the presidency. Indeed, since the first direct presidential election in 1965, both parties had dominated the Fifth Republic, alternating in power. Center-right president, Valéry Giscard d Estaing, is the outlier, but he enjoyed support from Gaullist Jacques Chirac in In previous presidential elections, the total score of these two parties had never gone below 36% and averaged above 50%. They barely reached 26% this time (see Fig. 1). It was no surprise that the socialist candidate, Benoît Hamon, did not reach the second round; he had been behind in the polls for a long time. However, his score was a catastrophe. Hamon lost most of the socialist core voters to Mélenchon and Macron. On the one hand, Mélenchon crushed him in both former socialist, lower-class, urban strongholds where Hollande had done much better at the previous election (La Courneuve: 10.3% against 44.4% for Mélenchon; Hollande 47% in 2012) and former rural strongholds (Ariège: 7.9% against 26.8% for Mélenchon; Hollande 34.4% in 2012). On the other hand, Macron beat Hamon in former socialist, upper-class, urban

5 A new party system in the making? The 2017 French Socialist candidate Gaullist candidate Fig. 1 Score of Gaullist and Socialist candidates in the first round of presidential elections ( ). Source: Constitutional Council. Note: Socialist candidate François Mitterrand was endorsed by the Communist Party in 1965 and 1974 strongholds (the 3rd district of Paris: 10.6% against 45% for Macron; Hollande 40.1% in 2012). Political parties rarely die. But when they do, they lose their core voters first (Mack 2010). Here is one clue that the Socialist party might actually disappear, but further investigation is needed before drawing a definitive conclusion. Marine Le Pen obtained the best ever result for an extreme-right candidate at a French presidential election (21.3%). It was the second time that the National Front reached the runoff. When Jean-Marie Le Pen pushed former Prime Minister Lionel Jospin out of the race in 2002, the result was met with incredulity (Cole 2002; Kuhn 2002; Miguet 2002). This was not the case this time. The National Front made a big breakthrough at the 2014 European elections (24.9%), and this result was confirmed by both the 2015 departmental (25.2%) and regional elections (28%). For this reason, Marine Le Pen was expected to lead the polls. In this context, her score was in fact below forecasts. The second round was not close: Emmanuel Macron defeated Marine Le Pen with 66.1% of the vote to become the eighth President of the Fifth Republic. This was the second-largest margin of victory in a presidential election, just behind Jacques Chirac s landslide win over Jean-Marie Le Pen in Many observers considered that Marine Le Pen s poor performance in the last presidential debate four days before the runoff was crucial to explaining her disappointing result. This type of reasoning overestimates short-term factors, however. Marine Le Pen had already underperformed in the first round. Part of the explanation can be traced back to issue salience during the campaign. Immigration has never been a top priority and Le Pen herself chose to focus on economic issues, as other contenders did. 1 1 Recent work has demonstrated that French mainstream parties have been eager to take up issues associated with niche parties (Meguid 2008; Brouard et al. 2012; Persico 2015), which is what Sarkozy

6 308 F. Gougou, S. Persico Le Pen s defeat was in line with recent second-order elections. Despite high-level support in the first round, the National Front still appears unable to break the glass ceiling at the second round (Labouret 2016). In the 2015 departmental elections, it was beaten in about 98% of runoffs (Shields 2015). In the 2015 regional elections, it failed to win a single region. When the stakes are high, the National Front continues to suffer from a front républicain, i.e., the willingness of French voters to keep the radical right out of power. In this context, as long as Macron was ahead in the first round, he was the favorite to take office. This is what happened. According to the polls, he was the rare case of a Condorcet winner. The structure of the French party system: new conflict dimensions and party blocks This section assesses the structure, the nature and the format of the new French political space. The political landscape after the 2017 presidential election looks quite different from the one before the election. The situation seems to point to a new pattern of party competition. Since the 1980s, French politics has experienced the parallel development of cleavages related to both globalization especially the politicization of immigration and the environmental crisis, which have cut across the traditional class and religious cleavages (Bornschier and Lachat 2009). This evolution has been common to most advanced industrial democracies: the politicization of new conflicts between libertarian and ethnocentric values, as well as between ecologists and productivists, has indeed transformed most West European polities (Inglehart 1984; Kitschelt 1994; Hooghe and Marks 2006; Kriesi et al. 2008, 2012). Yet, the way in which these cleavages have transformed party systems is different from one country to another, depending on the previous cleavage configuration. In France, it led to the rise of an extreme-right party, and, to a lesser extent, to the development of a Green party. It also changed the nature and content of party competition: while the cultural dimension of competition was mostly composed of religious issues, it was now structured by immigration and law and order (Tiberj 2012). In terms of the economic dimension of competition, issues linked to globalization overtook issues like privatization or state intervention in the economy (Bornschier and Lachat 2009). The environmental issue cut across the two main dimensions and had a cyclical impact on political agendas (Dalton 2009; Persico 2015). Overall, between the 1980s and the mid-2010s, the French party system was structured around one left-wing block, one moderate right-wing block and one extreme-right block what scholars called tripartition (Grunberg and Schweisguth 1997, 2003). Yet the difference between conservative and extreme-right voters seemed to fade away after the 2007 election, in relation to the radicalization of conservative voters regarding immigration and law and order (Gougou and Labouret 2013). In other Western countries, the Footnote 1 continued successfully did with immigration in 2007 (Sauger 2007; Spoon 2008). Yet, the 2017 election shows that mainstream parties will stop taking up those issues if niche parties no longer politicize them.

7 A new party system in the making? The 2017 French 309 development of a new tripolar party competition was witnessed as well (Deegan- Krause 2007; Bornschier 2010). Does the four-way battle in the first round of the French 2017 presidential election indicate that this standard tripartition is over? In order to answer this question, we map the 2017 French political space. This means assessing the conflict dimensions that play a role in structuring the party system and the role they play. This also means evaluating the connection between citizens preferences and values regarding these conflict dimensions and their vote on April 23, In other words, we will assess what issues differentiate one voter from one another and which value set leads to one candidate rather than another. This first step allows us to evaluate the format and content of the new French party system. Hypotheses Many assumptions can be made regarding the nature of the new French political space. It might be the case, as Le Pen puts it, that French politics is now mostly onedimensional, structured by the fight between patriots and globalists (Alduy 2017). It might be the case that the French political space is tripolar, as Macron suggests himself, but opposing progessist modernizers to both radical left and radical right challengers. Alternatively, Macron could simply be considered as a refurbished Socialist, in which case the Fifth Republic might soon come back to where it was before. Yet, the situation might also be more complex, with the fragmentation of the party system into four blocks representing new cleavage configurations. The solution to this puzzle depends on the answer to two distinct questions. 1. What does Macron s breakthrough mean? Are we witnessing the development of a new liberal centrist pole, independent from the old socialist and conservative parties, and which has never previously occurred under the Fifth Republic (Sauger 2004)? Alternatively, are we witnessing a reconfiguration of the socialist pole behind Macron, with the modernizers taking the lead over left-wing members? This generates our first two hypotheses. Hypothesis 1a Emmanuel Macron can be considered as leading a new liberal centrist pole, distinct from traditional parties. Hypothesis 1b Emmanuel Macron represents a new form of the traditional leftist pole. However, the fate of the French party system does not only depend on what Emmanuel Macron does. 2 Other players have a central role too. 2. What happens at the conservative end of the political spectrum? Has the division between the conservatives and the radical right remained untouched? Or is a new conservative-ethnocentric block being formed, rendering the historical difference between the moderate and extreme right obsolete? A 2 Let us not forget that the early days of the Fifth Republic proved how a new president with a new party could radically transform the political space.

8 310 F. Gougou, S. Persico supplementary set of hypotheses stems from the different answers to this second question. Hypothesis 2a rightist poles. François Fillon and Marine Le Pen still represent two distinct Hypothesis 2b François Fillon and Marine Le Pen now belong to the same rightist pole with only differences of degree between them. Table 2 represents the four main scenarios derived from the two sets of hypotheses. The first scenario four blocks indicates the greatest fragmentation, with Macron leading a new distinct centrist pole, Mélenchon and Hamon filling up most of what remains of the left-wing pole. In this scenario, Fillon and Le Pen still represent distinct rightist poles. This interpretation was defended by most commentators following the election and means that the former electoral patterns of the Fifth Republic have been changed for good. The second scenario standard tripolarity indicates stability vis-à-vis the former electoral order (Grunberg and Schweisguth 2003). Here, Macron represents the traditional leftist pole: he replaces the Socialist Party as the dominant left-wing party and shares this political space with Mélenchon and Hamon. The right remains divided between a moderate conservative pole represented by Fillon and an extreme-right pole represented by Le Pen. This configuration is in line with existing theories regarding party system change in Western Europe (Kriesi et al. 2012; Oesch and Rennwald 2017). The third scenario new tripolarity indicates a clear departure from the previous political order, with Macron leading a new independent centrist pole distinct from Melenchon and Hamon s traditional left while there is a fusion of the two rights with some of the more moderate right-wing leaders and voters joining Macron, as they have already started to do with the formation of the new government. Martin (2017) sketches this scenario with three poles: one democrateco-socialist; one liberal-globalizer and one conservative-identity pole. Finally, the fourth scenario quadrille bipolaire means that the Fifth Republic s political space reverts back to the situation prior to the 1980s with four parties of roughly equal strength, dividing the electorate into left and right coalitions (Duverger 1985). This means that the party system is now structured by a central conflict dimension, creating two blocks with an impassable frontier and varying relations of competition and cooperation between parties from the same block. Table 2 Hypotheses regarding the structure and form of the new French party system H2a Fillon and Le Pen as distinct rightist poles H2b Fillon and Le Pen on the same rightist pole H1a H1b Macron as a new distinct centrist pole Macron as a traditional leftist pole Four blocks Standard tripolarity New tripolarity Quadrille bipolaire

9 A new party system in the making? The 2017 French 311 Data and methods We use data from the French Election Study 2017, a cross-sectional post-electoral survey carried out after the second round of the presidential election (Gougou and Sauger 2017). Six main value dimensions were measured among respondents: authoritarianism (AUT), ecologism (ECO), ethnocentrism (ETH), economic globalization (GLO), economic liberalism (LIB) and social conservatism (MOR). Three items were used to capture each of these key value dimensions (Table 3). Respondents were asked to what extent they agree or disagree with particular statements; measurements relied on Likert-scale response formats from Strongly agree to Rather agree, Rather disagree and Strongly disagree. Following Table 3 Value orientations in the 2017 French Election Study In file In analysis Measurement Wording O3 AUT1 Categorical (2) School should: instill a sense of discipline/educate people for critical and open mindedness O9 AUT2 Ordinal (4) Death penalty should be restored O47 AUT3 Ordinal (4) In our society, one needs hierarchy and leaders O10 ETH1 Ordinal (4) Immigrants presence in France is a source of cultural enrichment O11 ETH2 Ordinal (4) French Muslims are French citizens like others O12 ETH3 Ordinal (4) There are too many immigrants in France O7 MOR1 Ordinal (4), Homosexuality is an acceptable way to express one s sexuality recoded (3) O8 MOR2 Ordinal (4) It is appropriate that same-sex couples may adopt children O46 MOR3 Ordinal (4), recoded (3) It is appropriate that a woman may abort O4 ECO1 Categorical (3), recoded (2) Climate change is a phenomenon: caused by human activity/natural, as Planet earth has known in the past/of which the origin remains unknown O6 ECO2 Ordinal (4) In order to really protect the environment for future generations, we will have to seriously reduce economic growth O64 ECO3 Ordinal (4) Are you very much in favor, in favor, opposed, or strongly opposed to energy production by nuclear plants? EU5 GLO1 Ordinal (5) Custom barriers and economic protectionism should be restored in France EU6 GLO2 Ordinal (5), recoded (4) Globalization clearly reduces national governments power over economic issues O49 GLO3 Ordinal (4) Economic consequences of globalization are extremely negative for France O1 LIB1 Categorical (2) In order to face economic difficulties, government should: trust private businesses/control private businesses more closely O2 LIB2 Categorical (2) In the next year, priority must be given to: increase French economy s competitiveness/better worker s situation O50 LIB3 Ordinal (4) In order to obtain social justice, one should take from the rich to give to the poor

10 312 F. Gougou, S. Persico standard conventions, we merged categories that did not include more than 5% of the respondents into Agree or Disagree statements. In order to map the 2017 French political space, we rely on geometric data analysis (GDA). GDA is a technique to understand the nature of data under investigation. The analysis allows us to reduce underlying structures into a small number of meaningful dimensions factors or axes (Le Roux and Rouanet 2010). In such a framework, the results are mainly descriptive: the technique detects relations between variables without making any assumption of causality. The primary aim is to let data speak for themselves : exploring relations between variables within a dataset is supposed to reveal patterns and structures of interest, whether expected or not. As value orientations in the dataset are categorical variables, we use multiple correspondence analysis (MCA). 3 MCA provides a geometric model of data, representing a set of individuals by a cloud of points. The statistical distance between two respondents to a survey is determined by their responses to the questions to which they gave different answers. The cloud of modalities is then derived from the cloud of individuals: it describes the distance between each answer to each question. Both clouds have exactly the same number of axes (dimensions) and the same eigenvalue (overall variance). The first axis is determined according to its eigenvalue and provides the best one-dimensional fit of the cloud, i.e., the maximal variance for one axis. Once the sequence of axes has been defined (with decreasing eigenvalues), the number of axes depends on the eigenvalues. Finally, the axes can be interpreted on the basis of the cloud of modalities and the statistical contribution of each category to the axis. Interpreting axes is the most challenging step in the analysis: following Benzécri (1992, p. 405), axes should express with conciseness and precision, the contrast (or opposition) between two extremes. Some additional variables can be put into the analysis without contributing to the determination of axes. These are called supplementary variables: their categories are plotted in the cloud of modalities and may help us to understand the relations between active variables. The mean point of each supplementary variable category can also be plotted in the cloud of individuals, along with concentration ellipses (which contain about 86% of the points of the sub-cloud). We chose to introduce two supplementary variables in our analysis, i.e., the vote in the first round of the 2017 presidential election, on the one hand, and issue saliency, on the other. This allows us to describe how the electorate fits into the current French political space. The new French political space The analysis is divided into three steps. First, we decide how many axes should be interpreted. Second, we visualize the cloud of categories and interpret the remaining axes. Third, we investigate the cloud of individuals and look at the concentration 3 We performed a specific multiple correspondence analysis (using Spad). This technique allows us to record non-responses as passive modalities : individuals with missing answers remain in the analysis, but missing values are ignored when determining distances between individuals (Le Roux and Rouanet 2004).

11 A new party system in the making? The 2017 French Produc vist pro-globaliza on neoliberal pole ECO3 -- LIB3 ++ GLO3 ++ Na onalist an -immigra on authoritarian pole ETH2 ++ MOR1 ++ FILLON -- Public debt LIB2 ++ GLO2 ++ GLO1 ++ MOR2 ++ MOR2 + LIB1 ++ AUT3 ++ GLO3 + ECO1 -- ECO3 - LIB3 + MACRON MOR3 ++ DUPONT-AIGNAN Immigra on AUT1 ++ ECO2 - ETH1 ++ ETH3 ++ ETH2 - ETH3 -- GLO2 -- Unemployment GLO1 + AUT2 ++ MOR1 - AUT2 -- ETH1 -- GLO1 -- Law and order AUT3 + AUT2 MOR ETH3 - MOR3 - ETH1 - LE PEN GLO3 -- Terrorism GLO2 + ECO1 ++ ETH2 - AUT1 -- ETH1 + MOR1 -- AUT2 + ETH3 + GL O2 - MOR2 -- GLO1 = MOR2 - ETH2 + ECO2 + AUT3 -- GLO1 - Inequali es HAMON ECO3 + LIB2 -- GLO3 - LIB3 - MÉLENCHON Global warming ECO2 LIB3 -- LIB1 -- ECO2 ++ ECO3 ++ Cosmopolitan pro-immigra on libertarian pole AUT Ecologist an -globaliza on interven onist pole 1.0 Fig. 2 Multiple correspondence analysis of voters values (active variables) and vote choices (illustrative variables) in plane 1 3 ellipses around the vote for each major candidate in the first round of the presidential election. Our results show that French politics is currently shaped by a bi-dimensional value space. MCA identifies three axes (see Appendix, Table 4). The first eigenvalue is predominant, accounting for a large amount of the variance. Its modified rate (50.6%) is about two times the modified rate of axis 2 (27.9%) and six times the modified rate of axis 3 (8.7%). Once the modified rates for the first three axes are cumulated, we reach 86.4%. Eigenvalues then decrease very slowly, and the differences are too small to be considered. Axis 2, however, is not a meaningful conflict dimension for our analysis: it sketches a horseshoe effect (also known as Guttman effect ), i.e., a parabolic cloud shape expressing the ordering of categories (Le Roux and Rouanet 2010). In other words, it opposes moderate answer modalities (rather agree/rather disagree) to radical answer modalities (strongly agree/strongly disagree). Such a statistical by-product is not an issue in MCA. Since MCA does not assume any structure in the data, bringing out a hierarchical structure inherent to questionnaires is one of the dimensions of interest in statistical terms. Axis 3, on the other hand, shows a meaningful conflict dimension for political scientists: this is the reason why we now consider Axes 1 and 3 only. Moving to the cloud of categories, plane 1 3 represents the current value space in the

12 314 F. Gougou, S. Persico French electorate (see Fig. 2). Although visually analyzing the graph provides a general idea of content of each axis, measuring the statistical contributions of categories is helpful to better interpret main dimensions. Categories contributing above average to Axis 1 or Axis 3 are found in Appendix (see Table 5). The political space at the 2017 presidential election is hence shaped by two cleavages. The most important conflict dimension, Axis 1, opposes anti-immigration authoritarian and socially conservative views to pro-immigration and culturally liberal opinions. Such a result fits perfectly the rise of a new cultural cleavage that has been observed in France (and other Western democracies) for decades. This cleavage is well described in the literature, even if there is no consensus as to how it should be labeled: integration/demarcation (Kriesi et al. 2006), open/closed society (Chiche et al. 2000), identity/cosmopolitism (Martin 2007), etc. While it forms the cultural dimension of party competition, this cleavage now represents the main (and largely so) division in French politics. The second conflict dimension, Axis 3, features neoliberal, pro-globalization, productivist opinions on the one hand, and interventionist, anti-globalization, ecologist stands on the other. Although it relates to the economy, the content of this conflict dimension seems to have changed over time: it differs from the formerly dominant class cleavage (Bartolini 2007). However, this cleavage has not yet been extensively described in the literature, and there are no cross-national studies to compare its content. From our perspective, it captures the reshaping of the economic dimension of party competition since the beginning of the Great Recession of 2008 in two ways: first, the voter s position vis-à-vis economic globalization plays a central role; second, ecologist values are now clearly aligned with and contribute to the economic dimension of competition; indeed, they are almost orthogonal to the cultural dimension. This result contradicts most work on the New Politics, which tends to argue that the environment is a component of the new cultural dimension (Hooghe and Marks 2006). Although it retains a weaker structuring power than the identity/cosmopolitism cleavage, this new economic dimension confronts pro-globalization, probusiness and anti-environmental citizens to citizens sharing eco-socialist values. We now turn to the test of our hypotheses by locating our illustrative variables in this bi-dimensional political space. Hypothesis 1a is confirmed: Macron leads a new pole, which is distinct from traditional parties. His mean point stands in the top-right corner, at a distance from other candidates mean points. Macron has managed to define a new quadrant in French politics, combining both pro-globalization, neoliberal economics and cultural openness. His electorate is strongly opposed to Marine Le Pen s electorate on the cultural dimension and is opposed to traditional left-wing electorates on the economic dimension as well. This is confirmed by analyzing issue saliency: Macron is close to public debt and far away from inequalities. As far as hypothesis 2 is concerned, the conclusion is not as clear as it might at first appear. Given the mean point of their electorate, Marine Le Pen and François Fillon still represent distinct rightist poles in the 2017 French political space: hypothesis 2a thus seems to be confirmed. Marine Le Pen s electorate is almost exclusively defined by the cultural axis of party competition, its mean point being close to immigration as the most salient issue in France today. By contrast, the economic axis of party competition is predominant for François Fillon s electorate.

13 A new party system in the making? The 2017 French 315 a Na onalist an -immigra on authoritarian pole 1.2 N = 337 Out : 10.4% Area : Produc vist pro-globaliza on neoliberal pole MELENCHON Cosmopolitan pro-immigra on libertarian pole Ecologist an -globaliza on interven onist pole b Produc vist pro-globaliza on neoliberal pole 1.2 N = 97 Out : 13.4% Area : 1.21 Na onalist an -immigra on authoritarian pole HAMON Cosmopolitan pro-immigra on libertarian pole Ecologist an -globaliza on interven onist pole Fig. 3 Concentration ellipses in the two-dimensional space of values of, a Mélenchon voters, b Hamon voters, c Macron voters, d Fillon voters, e Dupont-Aignan voters, f Le Pen voters

14 316 F. Gougou, S. Persico c Produc vist pro-globaliza on neoliberal pole Na onalist an -immigra on authoritarian pole 1.2 N = 353 Out : 13.0% Area : MACRON Cosmopolitan pro-immigra onlibertarian pole Ecologist an -globaliza on interven onist pole d Produc vist pro-globaliza on neoliberal pole Na onalist an -immigra on authoritarian pole 1.2 N = 210 Out : 13.3% Area : FILLON Cosmopolitan pro-immigra on libertarian pole Ecologist an -globaliza on interven onist pole Fig. 3 continued

15 A new party system in the making? The 2017 French 317 e Na onalist an -immigra on authoritarian pole 1.2 N = 56 Out : 7.1% Area : Produc vist pro-globaliza on neoliberal pole DUPONT-AIGNAN Cosmopolitan pro-immigra on libertarian pole Ecologist an -globaliza on interven onist pole f Na onalist an -immigra on authoritarian pole 1.2 N = 291 Out : 11.7% Area : Produc vist pro-globaliza on neoliberal pole LE PEN Cosmopolitan pro-immigra onlibertarian pole Ecologist an -globaliza on interven onist pole Fig. 3 continued Such results are consistent with one key aspect of the standard tripolarity configuration, i.e., the fact that both right-wing poles are conservative on different dimensions cultural for the radical right, economic for the moderate right. However, it must be acknowledged that the moderate right Fillon s voters seems to have moved right on the cultural axis. Even though Fillon s voters are less

16 318 F. Gougou, S. Persico conservative than Le Pen s voters, they are still much more conservative than Macron s, Hamon s and Mélenchon s voters. This has not always been the case: in 1997, there was no difference between center-right and socialist voters on the open/closed dimension (Chiche et al. 2000, p. 475). Exploring the cloud of individuals confirms our first results but allows us to go further (see Fig. 3a f). Ellipses of concentration around each electorate demonstrate that overlaps between candidates exist, but they remain marginal as far as the top four are concerned. Fillon, Le Pen, Macron and Mélenchon have very distinguishable positions in the quadrant, even when we take into account variations within their electorates. Macron s ellipse, however, is far larger than the others (area 1.76): managing to attract voters from all quadrants was crucial to explaining his strong performance in the first round of the presidential contest, and in the end, his election. Conclusion: the new French party system and the future of French electoral politics The 2017 French presidential election resulted in a fragmented political space. It was structured by one main conflict dimension immigration and authoritarianism and one less important one economy and environment leading to what could be seen as a new four-party system. The results presented in this article must obviously be put in a broader context: the presidential election was the first step in a sequence ending up with the legislative elections. In that regard, the four-block political space described in the first round of the presidential election has been slightly altered in the legislative election with the rise of La République en marche!, the new President s party, at the expense of the Republicans. Our main observations remain, however, unchanged. The first conflict dimension pertains to the cleavage between proponents of libertarianism and cultural openness from authoritarian and anti-immigrant citizens. Although this conflict dimension has structured French politics since the 1980s, it is now stronger than ever. On this dimension, Marine Le Pen and her voters take an extremely conservative position, which differentiates them from everyone else, especially Macron, Hamon and Mélenchon. Both Le Pen and Macron would benefit from an increased salience of this dimension in the political agenda. In this case, Macron would represent the left in the eyes of most left-wing voters, while Fillon s voters would continue leaning toward a conservative position. The second conflict dimension relates to the economy but cannot be subsumed into the old class cleavage. On this dimension, Mélenchon and Hamon clearly stand out with left-wing positions, while Macron s and Fillon s voters share neoliberal, pro-business and anti-environmental positions. The change in the economic dimension of competition opens up the space for the left-wing block: it clarifies the left s differences with Macron and makes pro-environmental positions largely shared by the electorate a clear component of the new economic divide. However, the organizational division between La France Insoumise, on the one hand, and the left-wing socialists and Greens, on the other hand, might be hard to overcome. In

17 A new party system in the making? The 2017 French 319 any case, Marine Le Pen would suffer from an increased visibility of this dimension, since her position is both unclear and controversial among her supporters. A last series of questions concerns the social-structural basis of these new cleavages. Which citizens tend to favor one block more than another? This article shows that the new cleavages have distinct organizational (different candidates and parties) and ideological (different value sets) bases (Bartolini and Mair 1990). Yet, it remains to be assessed what social groups compose each of the four blocks. The changing structure of the economy following the Great Recession and the changes in the middle-class way of life might provide answers to this set of questions (Oesch 2006; Bornschier 2007), but this has to be tested empirically. Appendix See Tables 4 and 5. Table 4 Multiple correspondence analysis: number of Axes to interpret Eigenvalues Inertia rate (%) Modified rate Cumulated rate (%) Axis Axis Axis Axis Axis Axis Axis Axis Axis Eigenvalues measure the variances of axes, and modified rates assess the importance of axes (Le Roux and Rouanet 1998). Three axes need to be interpreted Table 5 Multiple correspondence analysis: interpretation of Axes 1 and 3 Name Label Axis 1 Axis 3 Left Right Up Down AUT1?? School: discipline and effort AUT2?? Death penalty: strongly agree ETH1?? Immigration as cultural benefit: strongly disagree ETH3?? Too many immigrants in France: strongly agree MOR2?? Adoption by homosexuals: strongly disagree GLO1?? Protectionism: strongly agree AUT2- Death penalty: strongly disagree ETH1- Immigration as enrichment: strongly agree ETH2- French Muslims French as others: strongly agree 3.589

18 320 F. Gougou, S. Persico Table 5 continued Name Label Axis 1 Axis 3 References Left Right Up Down ETH3- Too many immigrants in France: strongly disagree GLO1- Protectionism: strongly disagree ECO2- Slow down economic growth: strongly disagree ECO3- Nuclear energy: strongly in favor LIB1?? Freedom and trust in firms LIB2?? Economic priority: competitiveness LIB3?? Take from the rich: strongly disagree GLO2?? Globalization on national economies: very bad GLO3?? Globalization on democracy: very bad LIB1- Control firms more closely LIB2- Economic priority: workers LIB3- Take from the rich: strongly agree Total Percentages indicate the relative contribution of each modality to axis according to its position on graph (cloud of modalities in plane 1 3) Alduy, C Ce qu ils disent vraiment: les politiques pris au mot. Paris: Éditions du Seuil. Bartolini, S The political mobilization of the European Left, : The class cleavage. Cambridge: Cambridge University Press. Bartolini, S., and P. Mair Identity, competition and electoral availability: The stabilisation of European Electorates, Cambridge: Cambridge University Press. Benzécri, J.-P Correspondence analysis handbook. New York: Marcel Dekker. Bornschier, S Social structure, collective identities, and patterns of conflict in party systems: Conceptualizing the formation and perpetuation of cleavages. In Politicising socio-cultural structures: Elite and mass perspectives on cleavages. Presented at the ECPR Joint Sessions Helsinki. Bornschier, S The new cultural divide and the two-dimensional political space in Western Europe. West European Politics 33 (3): Bornschier, S., and R. Lachat The evolution of the French political space and party system. West European Politics 32 (2): Brouard, S., E. Grossman, and I. Guinaudeau French party competition through the lens of electoral priorities. Revue française de science politique 62 (2): Chiche, J., B. Le Roux, P. Perrineau, and H. Rouanet L espace politique des électeurs français à la fin des années Nouveaux et anciens clivages, hétérogénéité des électoratsé. Revue française de science politique 50 (3): Cole, A A strange affair: The 2002 presidential and parliamentary elections in France. Government and Opposition 37 (3): Dalton, R.J Economics, environmentalism and party alignments: A note on partisan change in advanced industrial democracies. European Journal of Political Research 48 (2): Deegan-Krause, K New dimensions of political cleavage. In Oxford Handbook of Political Behavior, ed. R.J. Dalton, and H.-D. Klingemann, Oxford: Oxford University Press. Duverger, M Le système politique français, 18th ed. Paris: Presses Universitaires de France. Gougou, F., and S. Labouret Participation in the 2010 French regional elections: The minor impact of a change in the electoral calendar A reply to Fauvelle-Aymar. French Politics 9 (3):

19 A new party system in the making? The 2017 French 321 Gougou, F., and S. Labouret The end of tripartition? Revue française de science politique 63 (2): Gougou, F., and N. Sauger The 2017 French Election Study (FES 2017): A post-electoral crosssectional survey. French Politics 15(3). Grunberg, G., and E. Schweisguth Vers une tripartition de l espace politique. In L électeur a ses raisons, ed. D. Boy, and N. Mayer, Paris: Presses de Sciences Po. Grunberg, G., and E. Schweisguth La tripartition de l espace politique. In Le vote de tous les refus. Les élections présidentielles et législatives de 2002, ed. P. Perrineau and C. Ysmal, Paris: Presses de Sciences Po. Hooghe, L., and G. Marks Parties, public opinion, and identity: A postfunctionalist theory of European integration. Inglehart, R The changing structure of political cleavages in Western Society. In Electoral change in advanced industrial democracies: Realignment or dealignment?, ed. P.A. Beck, S.C. Flanagan, and R.J. Dalton, Princeton: Princeton University Press. Kitschelt, H The transformation of European social democracy. Cambridge: Cambridge University Press. Kriesi, H., E. Grande, M. Dolezal, M. Helbling, D. Höglinger, S. Hutter, and B. Wüest Political conflict in Western Europe. Cambridge: Cambridge University Press. Kriesi, H., E. Grande, R. Lachat, M. Dolezal, S. Bornschier, and T. Frey Globalization and the transformation of the national political space: Six European countries compared. European Journal of Political Research 45 (6): Kriesi, H., E. Grande, R. Lachat, M. Dolezal, S. Bornschier, and T. Frey West European politics in the age of globalization. Cambridge: Cambridge University Press. Kuhn, R The French presidential and parliamentary elections, Representation 39 (1): Labouret, S Le Front national au second tour en 2015: une force qui demeure impuissante. In Groupement de recherches sur l administration locale en Europe (Ed.), Les élections Locales Françaises Antony: Le Moniteur. Le Roux, B., and H. Rouanet Interpreting axes in MCA: Method of the contributions of points and deviations. In Visualization of categorical data, ed. J. Blasius, and M. Greenacre. San Diego: Academic Press. Le Roux, B., and H. Rouanet Geometric data analysis: From correspondence analysis to structured data analysis. Dordrecht: Kluwer. Le Roux, B., and H. Rouanet Multiple correspondence analysis. Los Angeles: Sage. Mack, C.S When political parties die: A cross-national analysis of disalignment and realignment, 1st ed. Santa Barbara, CA: Praeger. Martin, P Comment analyser les changements dans les systèmes partisans d Europe occidentale depuis 1945? Revue internationale de politique comparée 14 (2): Martin, P Un séisme politique. L élection présidentielle de Commentaire 158: Meguid, B.M Party competition between unequals: Strategies and electoral fortunes in Western Europe. Cambridge: Cambridge University Press. Miguet, A The French elections of 2002: After the earthquake, the Deluge. West European Politics 25 (4): Oesch, D Redrawing the class map: Stratification and Institutions in Britain, Germany, Sweden and Switzerland. New York: Palgrave Macmillan. Oesch, D., and L. Rennwald Electoral competition in Europe s new tripolar political space: class voting for the left, centre-right and radical right (EUI Working Paper No. 02). Florence: European University Institute. Persico, S Co-optation or avoidance? Revue française de science politique 65 (3): Sauger, N Entre survie, impasse et renouveau: les difficultés persistantes du centrisme français. Revue Française de Science Politique 54 (4): Sauger, N The French presidential and legislative elections of West European Politics 30 (5): Shields, J The Front National at the polls: Transformational elections or the status quo reaffirmed? French Politics 13 (4): Spoon, J.-J Presidential and legislative elections in France, April June Electoral Studies 27 (1): Tiberj, V Two-axis politics: Values, votes and sociological cleavages in France ( ). Revue française de science politique 62 (1):