Navigation – Plan du site

AccueilNumérosN°1, vol. 1EnquêtesTwo Lefts and Two Rights. Class Voti(…)

Two Lefts and Two Rights. Class Voting and Cultural Voting in the Netherlands, 2002

Précédé d'un avant-propos de Vincent Tiberj
Dick Houtman et Peter Achterberg

Résumés

This article elaborates and tests the so-called theory of the new political culture (Ronald Inglehart et al.) by means of the Dutch part of the European Social Survey (2002). The analysis is restricted on theoretical grounds to voting for parties representing new politics (centering on cultural issues: populist party (LPF, new right) versus green party (GroenLinks, new left) and old politics (centering on class issues: social-democratic party (PVDA, old left) versus conservative party (VVD, old right)). The class theory of politics explains voting for PVDA or VVD very well. The working class votes for the PVDA and the more privileged classes for the VVD, due to economic progressiveness and economic conservatism, respectively. A cultural logic underlies voting for Greens or LPF, however, with the well educated voting for the Greens and the poorly educated for the LPF, driven by libertarianism and authoritarianism, respectively. It is concluded that class voting needs to be carefully distinguished from cultural voting and that we may not so much have been witnessing a decline in class voting since World War II, as typically maintained, but rather an increase in cultural voting.

Cet article développe et teste la portée empirique de la théorie de la nouvelle culture politique (Ronald Inglehart et autres) en se fondant sur la partie néerlandaise de l’enquête sociale européenne de 2002 (European Social Survey 2002). L’analyse est restreinte pour les raisons théoriques au vote pour les partis emblématiques de cette nouvelle culture politique – parti populiste (LPF) de droite, d’une part, parti écologiste (GroenLinks), de l’autre – et aux deux grands partis traditionnels à base « classiste » – parti travailliste (PVDA) et parti conservateur (VVD). La théorie du vote de classe rend compte de manière satisfaisante du vote pour les partis traditionnels. La classe ouvrière vote pour le parti travailliste (PVDA) et les classes plus privilégiées pour le parti conservateur (VVD). Le vote pour le parti écologiste (GroenLinks) ou pour le parti populiste (LPF) semble relever davantage, en revanche, d’une logique de vote culturel, les catégories les plus éduquées portant davantage leurs suffrages sur les écologistes, et les moins instruits votant plus massivement pour le LPF sur la base de l’adhésion à des valeurs libérales/libertaires, d’un côté, et autoritaires, de l’autre. On en conclut à la nécessité de distinguer plus rigoureusement vote de classe et vote culturel, et que l’évolution marquante observée depuis la fin de la Deuxième Guerre mondiale n’est peut-être pas tant dans le déclin du vote de classe que dans la progression du vote culturel.

Haut de page

Texte intégral

The authors gratefully acknowledge Roy Kemmers for his invaluable research assistance and Philippe Coulangeon for his generous help and advice in finalizing the French text. A previous version of this paper has been presented by its first author at the colloque Culture et politique(s), 17 march 2007 in Lille (France).

Avant-propos

Vincent Tiberj

Chercheur/Senior Research Fellow
CEE-Sciences Po
vncent.tiberj@sciences-po.fr
27 rue Saint-Guillaume
75007 – Paris

1L’article qui suit est novateur à plus d’un titre. On peut déjà mettre en avant les résultats que les auteurs exposent sur la montée en puissance électorale des enjeux culturels sur un cas bien particulier, les Pays-Bas, longtemps considéré dans la littérature de politique comparée comme le pays européen le plus tolérant en matière d’immigration ou d’hédonisme. Ici se donnent à voir la « fin du consensus multiculturel » néerlandais (Sniderman, Hagendorn, 2007) et ses conséquences sur la montée de l’extrême droite dans les années 2000. Surtout ce qu’on peut et doit retenir tient au modèle proposé pour l’explication à la fois des évolutions du vote de classe et de la montée de la « nouvelle politique » (Kitschelt, 1995) et notamment celle de l’extrême droite. L’argument principal des auteurs a trait à l’explication de la fin de la « politique de classe » qui, loin d’avoir été remplacée par d’autres clivages sociaux, aurait pourtant perdurée, mais serait désormais « enterrée vivante » à la fois par la montée des enjeux culturels et les transformations de l’offre politique.

2Cet article permet ainsi de réinterpréter nombre des débats qui ont animé la sociologie électorale française dans les deux dernières décennies. Entendons-nous bien. Il ne s’agit pas ici de remettre en cause les analyses empiriques et les résultats obtenus par les équipes françaises mais plutôt de montrer combien cet article en désaxant le regard sociopolitique permet de mieux rendre compte des évolutions qu’on constate dans l’Hexagone.

3Il en va ainsi de l’évolution politique des classes populaires en général et des ouvriers en particulier. Il fut un temps où en France, comme dans nombre d’autres démocraties occidentales, le modèle explicatif le plus courant des comportements électoraux étaient celui du vote de classe (voir notamment Michelat et Simon, 1977), avec des niveaux remarquables de survote à gauche de la part des cols bleus. Ainsi, au 1er tour des présidentielles de 1974 et 1981 presque 2/3 des ouvriers choisissaient la gauche quand en 2002 et 2007, ils ne se distinguent plus du reste de la population (voir Gougou, 2007). Les raisons de ce mouvement ont été cherchées dans différentes directions, parmi lesquelles les plus marquantes sont la rupture d’avec la gauche de gouvernement et le changement social. Le divorce avec les classes populaires a été reproché à Lionel Jospin en 2002, mais cette critique peut être retracée jusqu’au choix de la rigueur de 1983 et le ralliement au marché qui s’ensuivit (Rey, 2004). Pour d’autres, on doit cette évolution aux mutations de la condition ouvrière (voir notamment Boy, Mayer, 1997) : précarisation et atomisation des conditions de travail, restructuration industrielle seraient autant de facteurs qui auraient érodé la « conscience de classe », l’élément structurant de la fidélité ouvrière à la gauche. Au-delà, on peut aussi noter l’idée de moyennisation de la condition ouvrière (Fourastié,  1979), qui aurait rendu moins prégnantes les inégalités sociales et  leurs traductions électorales, dans une interprétation que ne renierait pas le Ronald Inglehart des années 1980.

4Il en va aussi de l’irruption des enjeux culturels dans la sphère électorale. Déjà dans les années 1970 et avant, plusieurs sociologues des valeurs avaient noté cette tension face aux enjeux tels que l’immigration ou la libéralisation des mœurs (Girard et al., 1971). Mais cette tension ne va commencer à intéresser les spécialistes du vote qu’avec l’irruption du fn, notamment à partir de 1988. Ainsi, Gérard Grunberg et Étienne Schweisguth démontrent les tensions entre (anti)libéralisme économique et (anti)libéralisme culturel (Grunberg, Schweisguth, 1990) qui vont aboutir à la notion de tripartition électorale gauche/droite/extrême droite (Grunberg, Schweisguth, 1997) fondée à la fois sur les anciens enjeux gauche-droite (école, redistribution, rôle de l’État dans l’économie) et sur une nouvelle dimension autour de l’universalisme, l’immigration ou le libéralisme sexuel. Pierre  Martin démontre à rebours que cette tripartition est en fait devenue le nouvel ordre électoral caractérisant la compétition politique française dès 1984 (Martin, 2000).

5La question qui domine alors est celle de la primauté de la « nouvelle » dimension culturelle sur la dimension « traditionnelle » socioéconomique, dans un schéma explicatif se raccrochant à la révolution silencieuse postmatérialiste de Ronald Inglehart. C’est ainsi qu’à partir de l’enquête postélectorale de 1997, Jean Chiche et ses collègues caractérisent une dimension « ouverte/fermée » qui serait devenue « la dimension qui structure le plus les oppositions internes de l’électorat français […]. Ce n’est qu’ensuite qu’apparaissent les dimensions plus classiques du “social” et de l’“économie”, qui, depuis des décennies, façonnaient l’espace politique français en sous-tendant les oppositions gauche-droite » (Chiche et al., 2000, p. 470).

6Une fois brossé ce (bref) panorama empirique du cas français, on ne peut qu’être frappé par le modèle explicatif proposé par Achtenberg et Houtman, son principe de parcimonie et son potentiel d’application dans l’Hexagone. Plutôt que de jouer une dimension contre l’autre, les auteurs considèrent que les deux, la socioéconomique comme la culturelle, continuent de peser dans les urnes, à ceci près que l’offre politique et son évolution centripète sur la première laissent désormais plus de place à l’expression de la seconde. En cela, le culturel aurait « enterré vivant » les racines normatives (égalité de revenu et de redistribution) du vote de classe. Comment ne pas y voir un parallèle avec les études sociologiques du vote fn qui ont noté l’apparition dans les années 1990 des gaucho-lepénistes (Perrineau, 1995) ou des ouvriéro-lepénistes (Mayer, 2002), ces cols-bleus qui laissent d’abord s’exprimer leur rejet de l’immigration alors même qu’ils sont toujours en demande de redistribution et de plus d’égalité économique ? Les auteurs proposent également une typologie du système politique néerlandais mettant aux prises quatre pôles — ancienne gauche et nouvelle gauche, ancienne droite et nouvelle droite —, comment ne pas de nouveau être saisi par la similarité avec les résultats empiriques appuyant la tripartition de l’offre électorale française ? À ceci près que la nouvelle gauche française paraît moins puissante que son homologue néerlandaise.

7Reste que le dialogue entre ces deux cas empiriques peut aussi aller à rebours. Ainsi, la dernière élection présidentielle française de 2007 pourrait marquer une rupture sociopolitique, autre que celle proclamée par celui qui l’a emportée. D’abord, elle a vu une décroissance de l’influence du fn (Mayer, 2007) qui semble se poursuivre depuis. Ensuite, l’élection de Nicolas Sarkozy ne marque pas une « droitisation de l’électorat », bien au contraire, puisque jamais celui-ci n’a été aussi marqué par le libéralisme culturel et n’avait pas depuis longtemps été aussi en demande de gauche traditionnelle (Schweisguth 2007 ; Tiberj, 2008). Ce qui a permis la victoire de la droite tient à ce que son candidat a su faire la synthèse entre ancienne et nouvelle droite, entre conservatisme sociétal et libéralisme économique, permettant, pour l’heure, de l’emporter. En cela, on peut se demander si la contradiction entre les deux dimensions de valeurs n’est pas en phase de résolution en France entre un pôle social-libertaire incarné par la gauche et ce nouveau pôle réunifié à droite (Gougou, Tiberj, à paraître). Est-ce parce que le cas français a connu une extrême droite puissante bien avant les Pays-Bas, ou bien doit-on y voir un effet des modes de scrutin ? Quelle incidence cette « nouvelle donne » aura-t-elle sur la politique de classe et les prochains alignements électoraux ? Voilà un agenda de recherche pour les années à venir.

Boy D., Mayer N. (1997), « Que reste-t-il des variables lourdes ? », in Boy D., Mayer N. (dir.), L’électeur a ses raisons, Paris, Presses de Sciences Po, 1997, p. 101-138.

Chiche J., Le Roux B., Perrineau P., Rouanet H. (2000), « L’espace politique des électeurs français à la fin des années 1990 », Revue française de science politique, 50(3), juin, p. 463-487.

Flanagan S. & InglehartR. (1987), « Value change in industrial societies », American political science review, vol. 81, no 4, décembre, p. 1289-1319.

Fourastié J. (1979), les Trente glorieuses ou la révolution invisible de 1946 à 1975, Paris, Fayard.

Girard A. et al. (1971), « Attitudes des Français à l’égard de l’immigration étrangère », Population, no 5, p. 827-863.

Gougou F. (2007), « Les mutations du vote ouvrier sous la Ve république », Nouvelles Fondations, no 5, p. 15-20.

Gougou F., Tiberj V. (à paraître), « Les électeurs du Parti socialiste : sociologie de l’électorat d’un parti en reconversion (1981-2007) », in Lefebvre R. et Sawicki (dir.).

Grunberg G., Schweisguth E. (1997), « Vers une tripartition de l’espace politique », in Boy D., Mayer N. (dir.), L’électeur a ses raisons, Paris, Presses de Sciences Po, p. 179-219.

Grunberg G., Schweisguth E. (1990), « Libéralisme culturel, libéralisme économique », in cevipof L’électeur français en questions, Paris, Presses de Sciences Po, p. 45-69.

KitscheltH. (1995), The Radical Right in Western Europe: a Comparative Analysis, Ann Arbor, University of Michigan Press.

Martin P. (2000), Comprendre les évolutions électorales : la théorie des réalignements revisités, Paris, Presses de Sciences Po.

Mayer N. (2002), Ces Français qui votent Le Pen¸Paris, Flammarion.

Mayer N. (2007), « Comment Nicolas Sarkozy a rétréci l’électorat, Le Pen », Revue française de science politique, 57(3-4), juin-août, p. 429-445.

Michelat G. & Simon M. (1977), Classe, religion et comportement politique, Paris, Presses de Sciences Po/ Éditions sociales.

Perrineau P. (1995), « La dynamique du vote, Le Pen : le poids du gaucho-lepénisme », in Perrineau P., Ysmal C. (dir.), Le Vote de crise. L’élection présidentielle de 1995, Paris, Presses de Sciences Po, p. 243-262.

Rey H. (2004), la Gauche et les classes populaires : histoire et actualité d’une mésentente, Paris, La Découverte.

Schweisguth E. (2007), « Le trompe-l’œil de la droitisation », Revue française de science politique, 57(3), juin, p. 393-410.

SnidermanP., HagendoornL. (2007), When Ways of Life Collide: Multiculturalism and Its Discontents in the Netherlands, Princeton, Princeton University Press.

Tiberj V. (2008), La Crispation hexagonale : France fermée contre France plurielle 2001-2007, Paris, Plon.

New politics as new-leftist politics?

8Ever since his book The Silent Revolution (1977) American political scientist Ronald Inglehart has maintained that due to increasing affluence, “postmaterialist” values, pertaining to the primacy of individual liberty and self-attainment, have moved center stage in Western countries. This is held to have resulted in the emergence of a “new political culture”, that has increasingly overshadowed the “old political culture”, central to which were issues concerning the distribution of wealth and income between society’s classes (Inglehart, 1977, 1990, 1997; see also: Dalton et al., 1984; Rempel and Clark, 1997; Clark, 1998, 2001; Hechter, 2004).

  • 1 . With the founding of ppr and psp the Netherlands witnessed the emergence of two new-leftist parti (...)

9This new political culture has not left the political landscape unaffected either, given the emergence of new left-libertarian parties such as Les Verts in France, Die Grünen in Germany, and GroenLinks in the Netherlands (Inglehart, 1990, p. 281-283; see also: Hoffman-Martinot, 1991)1. Indeed, ever since the appearance of Inglehart’s pathbreaking book (1977), the idea that “there are now two Lefts, (…) which are rooted in different classes” (Lipset, 1981, p. 510) has been a mainstay in political sociology and political science alike (see also: Weakliem, 1991). These new left-libertarian parties are held to attract well-educated voters in particular, because these are more than others characterized by postmaterialism. These parties are thus held to play a major role in undermining the familiar alignment of the middle-class with the right (Inglehart, 1997, p. 254).

  • 2 . Examples of electorally successful new rightist-populist parties in Europe are the fpö in Austria (...)

10Inglehart’s influential theory has not remained uncontested for long, however. Its critics have pointed out that it is in fact quite problematical to exclude the possibility of “rightist-authoritarian postmaterialism” by definitional fiat (Flanagan, 1979, 1982, 1987; Middendorp, 1991, p. 262; Dekker et al., 1999). This objection makes all the more sense, because new rightist-populist parties have been electorally successful all over Europe since the 1980’s, constituting a veritable “silent counter-revolution” (Ignazi, 1992) that flies in the face of Inglehart’s claim that only left-libertarian issues are central in the new political culture (see also: Ignazi, 2003; Veugelers, 2000)2.

11What Inglehart’s critics argue, is that the theory of the new political culture needs to be broadened so as to acknowledge the existence of its rightist-authoritarian branch, too. In making this argument, they assume that voting for new-rightist parties can be explained through basically the same mechanism as voting for new-leftist ones —although working in the reverse direction, of course. It is assumed that in both cases voting is driven by “new” cultural issues that as such tend to reverse the traditional class-party alignments as these emerge from the “old” class-based economic interests. We study in the current paper whether Inglehart’s critics are right in this and to do so, we start with a critical discussion of how so-called “class voting” has been studied in the past and why this conventional approach mixes it up with the allegedly “new” type of “cultural voting” rather than systematically disentangling the two types (see also Houtman, 2001, 2003; Achterberg, 2006; Achterberg & Houtman, 2006).

Class voting and cultural voting: hypotheses

Class voting and its measurement: The Alford Index

12Changes in levels of class voting are typically measured as changes in the tendency of the working class to vote for leftist parties and the middle-class to vote for rightist ones. This has become pretty much the standard procedure in the study of class voting since Robert Alford (1967, p. 80) proposed in the 1960’s what has since come to be known as the “Alford Index”, i.e., measuring class voting “by subtracting the percentage of persons in non-manual occupations voting for “Left” parties from the percentage of manual workers voting for such parties”. So, the more frequently workers vote for leftist parties and the less frequently non-workers do so, the higher the Alford Index, and the higher the level of class voting. If workers vote exactly as frequently for leftist parties as non-workers, the Alford Index is zero, there is no evidence of class voting, and the class people belong to does not in any way affect how they vote. Although statistically more advanced varieties have been proposed since (e.g., Hout et al., 1993), these do not affect Alford’s theoretical rationale and hence tend to produce “the same conclusions with respect to the ranking of the countries according to their levels of class voting, and according to the speed of declines in class voting” (Nieuwbeerta, 1996, p. 370). The still widespread use of Alford’s Index can be demonstrated, for instance, by its application in many of the contributions in the edited volumes The Breakdown of Class Politics (Clark & Lipset, 2001) and The End of Class Politics? (Evans, 1999).

  • 3 . Nevertheless, Inglehart (1997, p. 4, 43, 47) holds that the shift towards “postmaterialism” is so (...)

13The Alford Index and its contemporary offshoots are however deeply problematic —even to such an extent that their increase or decline across time can tell us basically nothing about changes in the degree to which class drives voting (Houtman, 2003, p. 103-120). This is because its entire neglect of voting motives obscures the extent to which a relationship of a particular strength between class and voting emerges from a class-based economic motive (i.e., working-class economic egalitarianism and middle-class aversion to economic redistribution) on the one hand and from a cross-cutting cultural motive as emphasized by Inglehart (postmaterialism) and his critics (authoritarianism) on the other. This is problematical because although such a cultural motive is surely empirically related to class, it does nonetheless not stem from class-based economic interests like the economic motive for voting does. Clearly, then, we need to go beyond the simple bivariate relationship between class and voting by, firstly, actually including the class-based economic motive for voting that the Alford Index merely assumes and by, secondly, adding the cross-cutting cultural motive for voting as emphasized by advocates of the emergence of a “new political culture” (i.e., Inglehart and his aforementioned critics alike). We elaborate these shortcomings of the Alford Index in what follows, so as to develop an analytical distinction between what we will call “class voting” and “cultural voting”.3

Class voting and cultural voting

14A large number of studies since the 1950’s has pointed out that two political value domains exist in western democracies, which are by and large independent of one another among the electorates at large. The first domain, referred to as “economic conservatism/progressiveness” in this paper, pertains to the degree to which one favors either laissez-faire liberalism or economic redistribution by the state. The second domain, referred to as “authoritarianism/libertarianism” in this paper, pertains to the degree to which one favors either protection of individual liberty or maintenance of social order. This third value domain has in the meantime been demonstrated to include Inglehart’s notion of postmaterialism as well (Flanagan, 1979, 1982, 1987; Middendorp, 1991, p. 262; Dekker et al., 1999; Houtman, 2003, p. 66-82). “Thus, the distinctive claim of the postmaterialist argument is not that a second ideological dimension exists, but that it is having an increasing impact on voting”, as Weakliem (1991, p. 1330) rightly notes5.

15Already hinted at by Lipset in the 1950’s (1959), the zero-relationship between economic conservatism/progressiveness and authoritarianism/libertarianism has been found ever since among the mass publics of Western countries (e.g., Mitchell, 1966; Kelly & Chambliss, 1966; O’Kane, 1970; Felling & Peters, 1986; De Witte, 1990; Fleishman, 1988; Middendorp, 1991; Scheepers et al., 1992; Olson & Carroll, 1992; Heath et al., 1994; Evans et al., 1996; Houtman, 2003). This means that without additional information (Achterberg and Houtman, 2009), one cannot predict among the public at large whether, for instance, someone is for or against the death penalty (authoritarianism/libertarianism) if one knows his or her stance towards income redistribution (economic conservatism/progressiveness) (and vice versa, of course).

16Studies into the relationship between class and political values have found over and over again since World War II that authoritarianism does not emerge from a low income, but rather from a low-level of education (see Houtman, 2003 for a review of the relevant studies). The latter relationship has nevertheless often been interpreted as indicating that authoritarianism, just like economic progressiveness, emerges from a weak economic position. Likewise, the well-known positive relationship between education and postmaterialism has been interpreted by Inglehart (1977, 1990) as supporting his theory that growing up under conditions of affluence produces a long-lasting commitment to individual liberty and self-attainment: For a detailed theoretical discussion and empirical critique of these “Marxist lite” theories about authoritarianism and postmaterialism, the reader is referred to a previous book by one of us (Houtman, 2003), which demonstrates that both interpretations of this educational effect are flawed: authoritarianism and postmaterialism are neither driven by one’s present economic class position, nor by parental affluence during one’s formative years, but are instead intimately bound up with the amount of cultural capital one has at one’s disposal (compare Kalmijn, 1994; de Graaf & Kalmijn, 2001).

17The circumstance that two separate political value domains exist and that authoritarianism/libertarianism, unlike economic conservatism/progressiveness, does not so much emerge from class in an economic sense, but rather from cultural capital, calls for a new conceptualization of the relationship between class and voting. Although there is ample reason to be skeptical about the alleged economic basis of “postmaterialist” values (Houtman, 2003), this new conceptualization nevertheless gives due to Inglehart’s suggestion that voting is driven by two radically different logics. “(The) relation between class and vote cannot be described adequately by a single dimension” as Weakliem (1991, p. 1355) rightly notes, adding that “the conventional model of class politics may have been inaccurate even before it attracted widespread challenge.” Besides emphasizing the decisive role of cultural capital rather than class in an economic sense as the driving force behind cultural voting, the new conceptualization also refines Inglehart’s distinction between “class voting” and “voting according to postmaterialist values” (e.g., 1987, p. 1298). It acknowledges that class voting also entails “voting according to values”, although these are obviously radically different ones than those involved in cultural voting, and it replaces too narrowly defined “postmaterialism” by the more general “authoritarian/libertarian” value domain as the driving force behind cultural voting (for a similar conceptualization, see Middendorp, 1991).

18Figure 1 displays the new conceptualization of the relationship between class and voting. Class voting is now conceptualized as voting for a leftist party on the grounds of economically progressive political values generated by a weak class position (or, reversely, voting for a rightist political party on the grounds of economically conservative political values generated by a strong class position). Class voting as such needs to be distinguished from cultural voting, which is not driven by class-based economic interests, but rather by cultural capital and related authoritarianism/libertarianism. Cultural voting, then, is voting for a leftist political party on the grounds of libertarian values generated by ample cultural capital (or, reversely, voting for a rightist political party on the grounds of authoritarian political values generated by limited cultural capital).

Figure 1. Distinguishing class voting (path 2 x path 3) from cultural voting (path 4 x path 5).

Figure 1. Distinguishing class voting (path 2 x path 3) from cultural voting (path 4 x path 5).

19Given this distinction between class voting and cultural voting, it is likely that voting for old-leftist and old-rightist parties needs to be understood as class voting, while voting for new-leftist and new-rightist ones alike rather stands out as cultural voting. This is so, because new-leftist and new-rightist parties are precisely held to differ from old-leftist and old-rightist ones, because the former emphasize cultural issues rather than issues of economic distribution —with new-leftist parties emphasizing the need to expand individual liberty and tolerance for cultural diversity and new-rightist ones emphasizing the need to maintain social order, of course.

New left and new right in the Netherlands

  • 4 . Although the rightist-extremist Centre Party (later on: Centre Democrats) had managed to mobilize (...)

20Due to the traditional Dutch way of dealing with political pluralism —the “politics of accommodation” (Lijphart, 1968), i.e., aiming for consensus and carefully avoiding conflicts between minority groups—, the electoral breakthrough of new-rightist politics occurred much later in the Netherlands than elsewhere in Europe. Taboos on new-rightist and ethnocentrist political discourse were still firmly in place in this country when the late Pim Fortuyn started successfully attacking them in 2001. While rightist-populist parties had until then been marginal in the Netherlands4, Fortuyn’s new-rightist Populist Party (lpf) won no less than 17% of the votes in the national elections of 2002, after having already collected no less than 35% of the votes in the local elections in Rotterdam in March of that same year. Fortuyn’s landslide election victory has had lasting consequences for the Dutch political culture and the Dutch political landscape. The politics of accommodation and the multiculturalist discourse of the past have become increasingly contested since, resulting in a sharply polarized cultural and political climate and electoral successes of rightist-populist politicians (Houtman et al., 2008b; Houtman & Duyvendak, 2009).

21Because the 2002 national elections finally witnessed the historical breakthrough of new-rightist politics in the Netherlands, they offer the perfect opportunity to test our theory about the shortcomings of the Alford Index due to its mixing up of class voting and cross-cutting cultural voting. The new-rightist Populist Party (lpf), just like the new-leftist Greens (GroenLinks), does after all not so much present itself through issues concerning economic distribution, but rather through cultural ones. Whereas the Greens (GroenLinks) strongly emphasize the value of individual liberty and hence the rights of traditionally excluded cultural minority groups, the Populist Party (lpf) instead emphasizes the necessity of maintaining social order in the nation, especially by means of strict cultural assimilation of Muslim migrants. Both parties thus differ from the Labor Party (pvda) and the Conservative Party (vvd), the two large parties that have in the Netherlands traditionally represented the economic interests of the working class and the (upper-) middle-class, respectively.

22The assumption that “new” political parties are principally engaged in cultural politics leads to the hypothesis that voting for the new-rightist Populist Party (lpf) rather than the new-leftist Greens (GroenLinks) during the Dutch parliamentary elections in 2002 constitutes cultural voting rather than class voting. Voting for the old-rightist Conservative Party (vvd) rather than the old-leftist Labor Party (pvda), on the other hand, is expected to constitute class voting rather than cultural voting. If these hypotheses are confirmed, this suggests that we can indeed discern not only “two lefts”, but “two rights”, too. We hence test two additional hypotheses to find out whether such is the case. Firstly, Inglehart’s arguments suggest that voting for the new-leftist Greens (GroenLinks) rather than the old-leftist Labor Party (pvda) constitutes cultural voting by those with ample cultural capital. Secondly, we expect that voting for the new-rightist Populist Party (lpf) rather than the old-rightist Conservative Party (vvd) constitutes cultural voting by those with limited cultural capital.

Data and method

Data

23To test our four hypotheses, we analyze the data of the European Social Survey (2002) for the Netherlands, which contain relevant information pertaining to work, income, education, voting behavior, and opinions on all sorts of moral and political issues. The Dutch fieldwork for this international survey has taken place in the period between September 2002 and February 2003, hence shortly after the historical elections of May 2002. The sample size of 2,364 (based on a response rate of 68%) enables us to restrict ourselves to those who voted for the four aforementioned political parties, while still retaining a sufficient number of respondents. This restriction leaves us with 990 respondents, fairly distributed across the four parties: the new-leftist Greens (GroenLinks, 14%), the new-rightist Populist Party (lpf, 27%), the old-leftist Labor Party (pvda, 32%), and the old-rightist Conservative Party (vvd, 27%).

24These percentages match the shares won by the four parties in the actual elections fairly well: Greens (GroenLinks, 13%), Populist Party (lpf, 31%), Labor Party (pvda, 28%), and Conservative Party (vvd, 28%), adding up to 100% and accounting for 55% of the number of actually cast votes during the elections. The Christian Democrats were the largest of the six parties that we exclude from our analysis (28% of the votes in the actual elections) and none of the five other parties was able to collect more than 6% of the votes.

Measurement

  • 5 . We relied on either present occupation or last occupation (if retired or unemployed).

25Class —We use the so-called egp-class schema, designed by Erikson, Goldthorpe, and Portocarero, which assigns class positions to respondents on the basis of occupational title, self- employed status, and number of employees supervised (Erikson et al., 1979; Goldthorpe, 1980, p. 39-42). We have relied on the coding procedure developed by Ganzeboom and Treiman (2005)5.

26Education —Following Kalmijn (1994), De Graaf and Kalmijn (2001) and Achterberg and Houtman (2006) we measure cultural capital as level of education, distinguishing six educational categories: 1) No more than elementary education; 2) Lower vocational education (lbo, vmbo) or four-year secondary education (mulo, mavo); 3) Intermediary vocational education (mbo) or five-year secondary education (havo); 4) Pre-university education (hbs, vwo, Gymnasium); 5) Higher vocational education (ba); 6) University education (ma).

27Authoritarianism/libertarianism could be measured with eight items: “Important to do what is told and follow rules”, “Important to behave properly”, “Important to follow traditions and customs” (all three with six response categories), “Better for a country if almost everyone shares customs and traditions”, “Gays and lesbians free to live as they wish” (both with five response categories), “Country’s cultural life undermined or enriched by immigrants”, “Immigrants make country worse or better place to live”, and “Immigrants make countries” crime problems worse or better” (all three with five response categories). Principal component analysis produces a first factor with seven factor loadings between 0.45 and 0.67 and one (the one for the item on gays and lesbians) of 0.28. This first factor explains somewhat more than 30% of the variance: certainly not spectacular, but enough to construct a modestly reliable scale (Cronbach’s  = 0.64). Scores have been assigned as the means of the standardized items.

28Economic conservatism/progressiveness could unfortunately be measured with no more than two Likert-type items: “The government should take measures to reduce income differences” and “Employees need strong trade unions to protect their working conditions and wages” (both with five response categories indicating the degree of (dis)agreement). The correlation between the answers to the two questions is 0.30 and a principal component analysis produces a first factor that explains no less than 65% of the variance. Although this is substantial, with only two items it is impossible to construct a reliable scale (Cronbach’s  remains limited to 0.46). Scores have been assigned as the means of the two standardized items.

Hypotheses and statistical method

  • 6 . As a log odds ratio is the natural logs of the odds ratio, these estimates can be conversed into (...)

29Our theory holds, firstly, that choosing between the old-leftist Labor Party (pvda) and the old-rightist Conservative Party (vvd) constitutes class voting, while choosing between the new-rightist Populist Party (lpf) and the new-leftist Greens (GroenLinks) entails cultural voting (we test these hypotheses in section 4.1). Because new politics is not so much assumed to replace old economic issues with new cultural ones, but rather to add the latter to the former, our theory furthermore predicts that choosing between the new-leftist Greens (GroenLinks) and the old-leftist Labor Party (pvda), just like choosing between the new-rightist Populist Party (lpf) and the old-rightist Conservative Party (vvd), also entails cultural voting (we test these hypotheses in section 4.2). The two remaining possible comparisons between pairs of parties —i.e., between the old-rightist Conservative Party (vvd) and the new-leftist Greens (GroenLinks) on the one hand and between the old-leftist Labor Party (pvda) and the new-rightist Populist Party (lpf)— are meaningless for testing our theory. We therefore exclude these two comparisons from our analysis by not using multinomial logit modelling (which would after all produce estimations for all six possible comparisons between the four parties), but by instead relying on four separate binary logistic regressions for the four theoretically meaningful pairs of parties only. In the subsequent analyses, we estimate log-odds ratios —showing the natural log of the odds of voting for a party as the ratio of the probability of voting for one party over the probability of voting for the other6. Higher log odds ratios therefore mean that the probability to vote for one party outweighs the probability of voting for the other party. For negatively signed log-odds ratios the reversed is true.

30Because our aim is to expose the weakness of the conventional measurement of class voting, we consider it wise to proceed carefully so as to prevent throwing out the baby with the bath water. We therefore test our ideas about cultural voting in the strictest and most conservative manner possible by biasing our analysis against it and hence biasing it in favor of finding strong class effects. We do so by entering education as a pseudo-interval variable with six categories rather than as a set of five dummies, hence allowing for linear effects of education only. Naturally, this reduces education’s explanatory power in comparison to the alternative of dropping this assumption of linearity by modelling it as a series of dummy variables. We model egp-class, on the other hand, as a series of six dummy variables. Another reason for doing so is of course that the seven egp-classes, unlike the six educational categories, do not simply constitute a hierarchy. Although egp-class and education are obviously quite strongly related —indeed, this is the very reason why it is necessary to control class effects for education effects and vice versa—, the relationship between the two is not at all so strong so as to make it impossible to statistically disentangle the two (see for details: Houtman 2003, p. 24-46).

Results

Class voting and cultural voting?

31Does voting for the “old” political parties constitute class voting, while that for the “new” ones constitutes cultural voting? Table 1 demonstrates that voting for either the old-leftist Labor Party (pvda) or the old-rightist Conservative Party (vvd) can indeed be conceived of as class voting. The odds of voting for the Labor Party (pvda) are highest for those with the most precarious economic positions —skilled and unskilled manual workers as well as the poorly educated—, after all, while those with more favorable economic positions tend to vote for the Conservative Party (vvd). This difference in voting behavior is moreover caused by stronger desires for economic redistribution among the former as compared to the latter, precisely as the class theory of politics predicts. It is abundantly clear, in short, that voting for the old-leftist Labor Party (pvda) and the old-rightist Conservative Party (vvd) can be characterized as class voting: class-based economic interests are decisive here.

Table 1. Voting for the old-rightist Conservative Party (VVD) or the old-leftist Labor Party (PVDA) explained (1 = Conservative Party, 2 = Labor Party, log-odds ratios with standard errors in parentheses, N = 580)

Independent variables

Model 1

Model 2

Model 3

Model 4

Higher professionals (I)1

--

--

--

--

Lower professionals (II)

0.20 (0.23)

-0.05 (0.26)

0.19 (0.24)

-0.05 (0.26)

Nonnmanual workers (III)

0.04 (0.30)

-0.18 (0.33)

0.02 (0.30)

-0.18 (0.33)

Petty bourgeoisie (IV)

-0.68 (0.48)

-0.94 (0.52)

-0.78 (0.49)

-1.03 (0.53)

Higher working class (V)

0.25 (0.45)

0.35 (0.54)

0.20 (0.46)

0.27 (0.54)

Skilled manual workers (VI)

1.25** (0.49)

0.93 (0.54)

1.32** (0.49)

1.01 (0.54)

Semi and unskilled manual workers (VII)

1.14** (0.36)

0.78 (0.40)

1.19** (0.37)

0.81* (0.40)

Education

-0.24* (0.11)

-0.04 (0.07)

-0.19** (0.06)

-0.19 (0.13)

Economic progressiveness

  

1.30*** (0.14)

  

1.31*** (0.15)

Libertarianism

  

  

0.46*** (0.13)

0.47** (0.15)

Constant

-0.03 (0.19)

0.40 (0.40)

-0.62 (0.48)

-1.17* (0.49)

Pseudo R2 (Nagelkerke)

0.10

0.32

0.13

0.34

1. Not included in analysis (reference category)

*p < 0.05; **p < 0.01; ***p < 0.001

32The picture changes dramatically when we attempt to explain votes for the new-leftist Greens (GroenLinks) and the new-rightist Populist Party (lpf), however (Table 2). Although the poorly educated tend to vote for the Populist Party (lpf) and the well educated for the Greens (GroenLinks), distinctions between egp classes have no explanatory value whatsoever in this case. And indeed, the votes for the Populist Party (lpf) by the poorly educated and for the Greens (GroenLinks) by the well educated are driven by high levels of authoritarianism and libertarianism, respectively, underscoring that education plays a cultural rather than an economic role here. Economic egalitarianism also leads to voting for the Greens (GroenLinks), to be sure, but its role is substantially weaker and the circumstance that it leads the well educated rather than the poorly educated to vote for a leftist party obviously flies in the face of the class theory of politics.

33These findings confirm, in short, that voting for the new-leftist Greens (GroenLinks) and the new-rightist Populist Party (lpf) indeed cannot be understood as class voting, but rather constitutes cultural voting. While it is perfectly able to explain votes for old-leftist and old-rightist parties, then, the class theory of politics has nothing to offer when it comes to the explanation of votes for new-leftist and new-rightist parties. To satisfactorily account for the latter, the cultural significance of education needs to be acknowledged, so that in this case education needs to be treated as an indicator for cultural capital rather than class in an economic sense. Whereas the old politics of class is defined by a conflict between the rich and the economically less well-off, in short, the new cultural politics pits a libertarian cultural elite against an authoritarian “cultural proletariat”.

Two lefts and two rights?

34Does a cultural gap exist between those who vote for the new-leftist Greens (GroenLinks) and those who vote for the old-leftist Labor Party (pvda), too?

Table 2. Voting for the new-rightist Populist Party (LPF) or the new-leftist Greens (GroenLinks) (1 = Populist Party, 2 = Greens, log-odds ratios with standard errors in parentheses, N = 392)

Independent variables

Model 1

Model 2

Model 3

Model 4

Higher professionals (I)1

--

--

--

--

Lower professionals (II)

-0.04 (0.31)

-0.06 (0.32)

0.03 (0.37)

-0.07 (0.39)

Nonnmanual workers (III)

0.58 (0.38)

0.45 (0.39)

0.61 (0.44)

0.46 (0.46)

Petty bourgeoisie (IV)

-0.30 (0.58)

-0.06 (0.60)

-0.25 (0.69)

0.06 (0.73)

Higher working class (V)

-0.30 (0.59)

-0.51 (0.60)

-0.44 (0.69)

-0.62 (0.70)

Skilled manual workers (VI)

-0.83 (0.75)

-1.25 (0.77)

-0.21 (0.80)

-0.54 (0.82)

Semi and unskilled manual workers (VII)

0.51 (0.41)

0.45 (0.43)

0.32 (0.47)

0.25 (0.50)

Education

0.71*** (0.15)

0.46*** (0.08)

0.24** (0.09)

0.59*** (0.18)

Economic progressiveness

  

0.77*** (0.16)

  

0.87*** (0.21)

Libertarianism

  

1.73*** (0.21)

1.75*** (0.14)

Constant

-0.89** (0.26)

-2.70*** (0.51)

-6.96*** (0.83)

-6.05*** (0.73)

Pseudo R2 (Nagelkerke)

0.11

0.20

0.42

0.48

1. Not included in analysis (reference category)

*p < 0.05; **p < 0.01; ***p < 0.001

  • 7 . In a previous publication we have already demonstrated that the same goes for voting for the new- (...)

35Table 3 points out that such is indeed the case. We do not so much find classes pitted against one another here, but rather the well educated cultural elite that tends to vote for the new-leftist Greens (GroenLinks) against the less well educated who tend to vote for the old-leftist Labor Party (pvda). Consistent with this, the highly educated are not driven by class-based economic reasons, confirming the assumption of the theory of the new political culture that voting for the new-leftist Greens (GroenLinks) rather than the old-leftist Labor Party (pvda) is culturally rather than economically driven7.

36The gap between the new-rightist Populist Party (lpf) and the old-rightist Conservative Party (vvd) is basically identical as that between the new-leftist Greens (GroenLinks) and those who vote for the old-leftist Labor Party (pvda). Again, we find a predominantly cultural gap, with not so much economic classes, but rather educational categories pitted against one another, and the poorly educated voting for the new-rightist Populist Party (lpf) rather than the old-rightist Conservative Party (vvd) because they are more authoritarian (Table 4).

37In striking contrast to what the virtual absence of literature about “two rights” suggests, the gap that we find here is moreover even wider than that between the new-leftist Greens (GroenLinks) and the old-leftist Labor Party (pvda). Variance explained is no less than twice as high as in the case of the old-leftist Labor Party (pvda) and the new-leftist Greens (GroenLinks). By the time of the Dutch parliamentary elections of 2002, in short, the emergence of a new political culture had led not only to a bifurcation between “new left” and “old left” in Dutch politics, but to a similar gap between “new right” and “old right” as well.

The Alford Index and its Great Disappearance Act

38Our findings point out that society’s lower strata do not necessarily vote for leftist parties, just like the more privileged classes do not necessarily vote for rightist ones. If the working class votes for leftist parties, it is driven by class-based economic interests; if it votes for rightist ones, cultural capital and cultural voting motives are decisive. The reverse applies to the (upper-) middle-classes: if they vote for rightist parties, this is based on their class-based economic interests, while if they vote for leftist ones, cultural capital and cultural voting motives are decisive (see also: Achterberg and Houtman, 2006). Class voting and cultural voting hence work in opposite directions and tend to cancel one another out in a way that the Alford Index misses and obscures. As a consequence, the Alford Index cannot be relied on to ascertain levels of class voting, because a bivariate relationship between class and voting does not so much capture class voting, but rather the extent to which the latter is stronger than reversed cultural voting, i.e., the net balance of class voting and cultural voting. In our case at hand here, the Alford Index obscures that class voting and cultural voting are both strongly present in the Netherlands nowadays. This is demonstrated in Table 5: although manual workers are somewhat more likely than others to vote for leftist parties in the Netherlands nowadays, the differences are only very slight with egp class explaining no more than a mere two percent of the differences in voting behavior. As we have seen, this does not mean that class voting is almost non-existent, as the traditional interpretation would be, but rather that class voting and cultural voting work in opposite directions.

Table 3. Voting for the old-leftist Labor Party (PVDA) or the new-leftist Greens (GroenLinks) (1 = Labor Party, 2 = Greens, log-odds ratios with standard errors in parentheses, N = 458)

Independent variables

Model 1

Model 2

Model 3

Model 4

Higher professionals (I)1

--

--

--

--

Lower professionals (II)

0.01 (0.29)

0.01 (0.29)

0.02 (0.30)

0.02 (0.30)

Nonnmanual workers (III)

0.53 (0.35)

0.53 (0.35)

0.51 (0.36)

0.51 (0.36)

Petty bourgeoisie (IV)

0.52 (0.62)

0.51 (0.62)

0.45 (0.63)

0.44 (0.63)

Higher working class (V)

-0.03 (0.59)

-0.04 (0.59)

-0.02 (0.60)

-0.03 (0.60)

Skilled manual workers (VI)

-1.22 (0.73)

-1.21 (0.73)

-1.09 (0.73)

-1.08 (0.74)

Semi and unskilled manual workers (VII)

0.13 (0.39)

0.13 (0.39)

0.14 (0.40)

0.14 (0.40)

Education

0.40** (0.13)

0.22** (0.07)

0.12 (0.08)

0.22 (0.14)

Economic progressiveness

  

-0.04 (0.16)

  

-0.04 (0.17)

Libertarianism

  

  

0.64*** (0.16)

0.64*** (0.16)

Constant

-0.96*** (0.25)

-1.80*** (0.45)

-3.46*** (0.64)

-2.99*** (0.59)

Pseudo R2 (Nagelkerke)

0.07

0.07

0.12

0.12

1. Not included in analysis

*p < 0.05; **p < 0.01; ***p < 0.001

Table 4. Voting for the old-rightist Conservative Party (VVD) or the new-rightist Populist Party (LPF) (1 = Conservative Party, 2 = Populist Party, log-odds ratios with standard errors in parentheses, N = 514)

Independent variables

Model 1

Model 2

Model 3

Model 4

Higher professionals (I)1

--

--

--

--

Lower professionals (II)

0.29 (0.25)

0.24 (0.25)

0.28 (0.26)

0.23 (0.26)

Nonnmanual workers (III)

0.07 (0.31)

-0.02 (0.32)

0.02 (0.33)

-0.07 (0.33)

Petty bourgeoisie (IV)

0.01 (0.43)

-0.02 (0.43)

0.03 (0.45)

-0.01 (0.45)

Higher working class (V)

0.49 (0.45)

0.35 (0.46)

0.58 (0.47)

0.45 (0.48)

Skilled manual workers (VI)

0.72 (0.52)

0.54 (0.53)

0.53 (0.53)

0.38 (0.54)

Semi / and unskilled manual workers (VII)

0.78* (0.38)

0.73 (0.38)

0.77* (0.39)

0.71 (0.39)

Education

-0.49*** (0.12)

-0.23** (0.07)

-0.23** (0.07)

-0.36** (0.13)

Economic progressiveness

  

0.30** (0.11)

  

0.28* (0.11)

Libertarianism

  

  

-0.80*** (0.14)

-0.79*** (0.14)

Constant

-0.13 (0.21)

0.92* (0.38)

2.95*** (0.53)

2.14*** (0.43)

Pseudo R2 (Nagelkerke)

0.10

0.12

0.19

0.24

1. Not included in analysis

*p < 0.05; **p < 0.01; ***p < 0.001

39In other words: although the traditional approach to class voting would lead to the conclusion that class voting is virtually non-existent in the Netherlands nowadays, it is in fact strongly present, but this presence is concealed by the Alford Index.

Table 5.
Voting for a rightist or a leftist party explained (1 = right, 2 = left, log-odds ratios with standard errors in parentheses, N = 972)

Independent variables

Model 1

Higher professionals (I)1

--

Lower professionals (II)

0.07 (0.18)

Nonnmanual workers (III)

0.14 (0.23)

Petty bourgeoisie (IV)

-0.47 (0.36)

Higher working class (V)

-0.01 (0.34)

Skilled manual workers (VI)

0.70* (0.33)

Semi and unskilled manual workers (VII)

0.71** (0.25)

Education

0.05 (0.04)

Constant

-0.53* (0.27)

Pseudo R2 (Nagelkerke)

0.02

1. Not included in analysis (reference category)

*p < 0.05; **p < 0.01; ***p < 0.001

Conclusion and debate

40Our findings point out that Inglehart’s critics are correct in arguing that voting for new-rightist parties can be explained through basically the same mechanism as voting for new-leftist parties. In both instances we are dealing with cultural voting, i.e., voting for a leftist party on the grounds of libertarian values generated by ample cultural capital, or, reversely, voting for a rightist party on the grounds of authoritarian values generated by limited cultural capital. Moreover, as we have seen, a cultural gap exists not only between those who vote for either of the two lefts (the new-leftist Greens (GroenLinks) and the old-leftist Labor Party (pvda)), but also between those who vote for either of the two rights (the new-rightist Populist Party (lpf) and the old-rightist Conservative Party (vvd)). Voting for new left rather than old left emerges from a high level of libertarianism, rooted in a large amount of cultural capital, while voting for new right rather than old right on the contrary emerges from a high level of authoritarianism, rooted in a limited amount of cultural capital.

41Distinguishing this cultural type of voting more systematically from class voting in future research requires replacing the Alford Index with a more valid approach, because the former underestimates class voting by mixing it up with reversed cultural voting. Carefully distinguishing the two is even more important, because they can easily vary independently of one another. So, whereas Inglehart (1997, p. 254) is by and large correct in arguing that “Postmaterialists come from middle-class backgrounds” and that “middle-class postmaterialists move left (…) and working-class materialists move to the right”, he may well be mistaken in suggesting that such a movement is necessarily “conducive to a decline in class voting”. Declining Alford indices may after all not so much indicate declines in class voting, but rather increases in cultural voting, so that the decline of the familiar class-party alignments since World War II (Nieuwbeerta, 1996) may not so much have been caused by a breakdown of class politics and a decline of class voting, but rather by a dramatic proliferation of cultural politics and an increase in cultural voting. For evidence that such is indeed the case, the reader is referred to Achterberg (2006), Houtman et al. (2008a) and Van der Waal et al. (2007).

42Our findings clearly invoke the need for cross-national and historical analyses that disentangle class voting and cultural voting more carefully. A decline or increase of the Alford Index may after all mean basically anything. When the Alford Index declines, for instance, that may indeed mean that class voting has declined, as the conventional interpretation suggests. It may however also mean that class voting has remained stable (while cultural voting has increased) or even that class voting has increased rather than decreased (while cultural voting has increased even more). It is hardly surprising, then, that Nieuwbeerta’s (1995) attempt to explain the decline and cross-national variation of the Alford Index since World War II has led him to reject, almost without exception, hypotheses derived from the class theory of politics about the role of socio-economic context variables such as the size of income differences, the living standard, the percentage of intergenerational class mobility, trade union density, the relative size of the working class, etcetera. Van der Waal et al. (2007) have meanwhile demonstrated that this explanatory impotence of the class approach of politics is indeed caused by the unacknowledged ambiguity of the Alford Index, which measures cultural voting at least as much as it does class voting.

43The drops in the Alford Index during the heydays of the student protests of the 1960’s and 1970’s, for instance, are more likely to indicate increases in cultural voting than decreases in class voting, because of the sharply increased salience of cultural issues pertaining to individual liberty and democracy back then. The election victory of De Gaulle in France shortly after the vehement student unrest in Paris in May 1968, for instance, seems to have been caused by support by less-educated workers for De Gaulle’s emphasis on restoring order one the one hand and well-educated young people expressing their support for the Left on the other (see also Inglehart 1977: 267-284). Although this produced a drop in the Alford Index, it does hence not seem to signify a decline in class voting, but rather an increase in cultural voting. This suggests that a re-analysis of the actual causes underlying sharp drops or increases in the Alford Index in the past may be quite illuminating from a theoretical point of view. Indeed, going even further back in history, it is likely that the election victory that brought the German national socialists to power in the 1930’s was also based on cultural voting. In other words: there seem to be no good reasons to believe that cultural voting is something entirely new that has only emerged as late at the end of the 1960’s (as Inglehart suggests), although it seems undeniable that it has become much stronger since then.

44Although, reacting to his critics, Inglehart is now acknowledging the existence of new-rightist political currents, he argues that “new rightist groups are a reaction against broader trends that are moving faster than these societies can assimilate them” and hence maintains confidently that they “do not represent the wave of the future” (1997, p. 251). There is ample room for skepticism here. Inglehart’s index for postmaterialism is after all hardly fit to detect “the wave of the future”, because it excludes the possibility of right-authoritarian postmaterialism by definition, thus enabling it to fly under its radar and remain undetected (Flanagan, 1979, 1982, 1987). Moreover, and underscoring the severity of this shortcoming, the political salience of right-authoritarian issues has in fact increased even more strongly in Western industrial societies since World War II than that of left-libertarian ones (Achterberg, 2006; Houtman et al., 2008a). Contrary to Inglehart’s position, then, there are good reasons to assume that right-authoritarian politics is here to stay and grow, and his critics seem correct in arguing that the theory of the new political culture needs to be broadened, so as to incorporate its rightist-authoritarian branch alongside the left-libertarian one.

45Our findings also point out that the bifurcation between studies into voting for new-leftist parties on the one hand and for new-rightist ones on the other is quite unfortunate. While the former are typically based on the theory of the new political culture, the latter tend to be based on the assumption that new-rightist parties are either “protest parties”, so that their voters are driven by political distrust and cynicism, or “anti-immigrant parties”, to the effect that their voters are driven by racist and ethnocentric appeals. Although it has been proposed that anti-immigrant parties can be distinguished from protest parties (e.g., Fennema, 1997), such a distinction proves hard to apply in practice. This is underscored by the circumstance that even during the very brief period 1994-1999 particular anti-immigrant parties seem to have changed into protest parties (Van der Brug & Fennema, 2003). Attempts to construct solid boundaries between these two types of parties thus seem artificial and hence produce unstable results.

46Indeed, political cynicism and ethnocentrism are strongly related among themselves and to authoritarianism, while all of these are strongly and negatively related to postmaterialism, making it hardly surprising that they all drive new-rightist voting. Elchardus (1996) has demonstrated convincingly that a linear combination of these variables very well explains new-rightist Vlaams Blok voting in Flanders, Belgium. Like new-leftist parties, then, new-rightist ones are not single-issue parties either. Like the former, they deal with cultural issues in quite a general sense, although they are the former’s mirror images, of course. Whereas the new-leftist parties emphasize the desirability of increasing individual liberty and tolerance for cultural diversity, the new-rightist ones emphasize the desirability of an orderly nation, conceiving of “the people” as a homogeneous and undivided whole and conceiving of politicians who refuse to take “the will of the people” seriously as traitors. New politics is a politics beyond class, in short —a cultural politics that focuses on issues of individual liberty, social order, and identity.

Haut de page

Bibliographie

Achterberg P. (2006), Considering Cultural Conflict: Class Politics and Cultural Politics in Western Societies, Maastricht, Shaker.

Achterberg P. & HoutmanD. (2006), “Why Do So Many People Vote “Unnaturally”? A Cultural Explanation for Voting Behaviour”, European Journal for Political Research, 45(1), p. 75-92.

AchterbergP. & HoutmanD. (2009), “Ideologically “Illogical”? Why Do the Lower-Educated Dutch Display so Little Value Coherence?”, Social Forces 87(3), p. 1649-1670.

AlfordR. R. (1967), “Class Voting in the Anglo-American Political Systems”, in Lipset S. M. & Rokkan S. (eds) Party Systems and Voter Alignments: Cross-National Perspectives, New York, Free Press, p. 67-93.

Clark T. N. (1998), “Assessing the New Political Culture by Comparing Cities Around the World”, in Clark T.N. & Hoffman-Martinot V. (eds) The New Political Culture, Boulder (co), Westview, p. 93-194.

Clark T. N. (2001), “The Debate over “Are Social Classes Dying?”” in Clark T. N. & Lipset S. M. (eds) The Breakdown of Class Politics: A Debate on Post-Industrial Stratification, Baltimore (md), Johns Hopkins University Press, p. 273-319.

DaltonR. J., FlanaganS. C. & BeckP. A. (1984), Electoral Change in Advanced Industrial Democracies: Realignment or Dealignment?, Princeton (nj), Princeton University Press.

De GraafP.M. & KalmijnM. (2001), “Trends in the Intergenerational Transmission of Cultural and Economic Status”, Acta Sociologica 44(1), p. 51-66.

DekkerP., EsterP. & Van den BroekA. (1999), “Fixing Left and Right: Value Orientations According to Middendorp and Inglehart”, in De Witte H. & Scheepers P. (eds) Ideology in the Low Countries, Trends, Models and Lacunae, Assen, Van Gorcum, p. 151-176.

ElchardusM. (1996), “Class, Cultural Re-Alignment and the Rise of the Populist Right”, in Erskine A. (ed.), Changing Europe: Some Aspects of Identity, Conflict and Social Justice, Aldershot, Avebury, p. 41-63.

EriksonR., GoldthorpeJ. H. & PortocareroL. (1979), “Intergenerational Class Mobility in Three Western European Societies: England, France and Sweden”, British Journal of Sociology 30(4), p. 415-441.

EvansG. (ed.) (1999), The End of Class Politics? Class Voting in Comparative Context, Oxford, Oxford University Press.

EvansG., HeathA. & Lalljee M. (1996), “Measuring Left Right and Libertarian Authoritarian Values in the British Electorate”, British Journal of Sociology, 47(1), p. 93-112.

FellingA. & PetersJ. (1986), “Conservatism: A Multidimensional Concept”, Netherlands’ Journal of Sociology 22(1), p. 36-60.

FennemaM. (1997), “Some Conceptual Issues and Problems in the Comparison of Anti-Immigrant Parties in Western Europe”, Party Politics, 3(4), p. 473-492.

Flanagan S. C. (1979), “Value Change and Partisan Change in Japan: The Silent Revolution Revisited”, Comparative Politics, 11(April), p. 253-278.

Flanagan S. C. (1982), “Changing Values in Advanced Industrial Societies: Inglehart’s Silent Revolution from the Perspective of Japanese Findings”, Comparative Political Studies, 14(4), p. 403-444.

Flanagan S. C. (1987), “Value Change in Industrial Societies: Reply to Inglehart”, American Political Science Review, 81(4), p. 1303-1319.

Fleishman J. A. (1988), “Attitude Organization in the General Public: Evidence for a Bidimensional Structure”, Social Forces, 67(1), p. 159-184.

GanzeboomH. B. G. & Treiman D. J. (2005) “International Stratification and Mobility File: Conversion Tools.” (http://www.fsw.vu.nl/~h.ganzeboom/ismf).

GoldthorpeJ. H. (1980), Social Mobility and Class Structure in Modern Britain, Oxford, Clarendon Press.

Heath A. F., EvansG. & Martin J. (1994), “The Measurement of Core Beliefs and Values: The Development of Balanced Socialist/Laissez-Faire and Libertarian/Authoritarian Scales”, British Journal of Political Science, 24(1), p. 115-132.

HechterM. (2004), “From Class to Culture”, American Journal of Sociology, 110(2), p. 400-445.

Hoffman-MartinotV. (1991), “Grüne and Verts: Two Faces of European Ecologism”, West European Politics, 14(1): 70-95.

Hout M., Brooks C. & Manza J. (1993), “The Persistence of Classes in Post-Industrial Societies”, International Sociology, 8(3), p. 259-278.

HoutmanD. (2001), “Class, Culture, and Conservatism: Reassessing Education as a Variable in Political Sociology”, in Clark T. N. & Lipset S. M. (eds), The Breakdown of Class Politics: A Debate on Post-Industrial Stratification, Baltimore (md), Johns Hopkins University Press, p. 161-195.

HoutmanD. (2003) Class and Politics in Contemporary Social Science: “Marxism Lite” and Its Blind Spot for Culture, Hawthorne (ny), Aldine  de Gruyter.

HoutmanD., AchterbergP. & DerksA., (2008a), Farewell to the Leftist Working Class, New Brunswick (nj), Transaction (forthcoming).

HoutmanD., AchterbergP. & Duyvendak, J. W. (2008b), “De verhitte politieke cultuur van een ontzuilde samenleving”, in B. Snels and N. Thijssen (eds), De grote kloof: Verhitte politiek in tijden van verwarring, Amsterdam, Boom, p. 61-80.

HoutmanD. & DuyvendakJ. W. (2009), “Boerka’s, boerkini’s en belastingcenten: Culturele en politieke polarisatie in een post-Christelijke samenleving”, in Raad voor Maatschappelijke Ontwikkeling (ed.), Polarisatie: Bedreigend en verrijkend, Amsterdam, swp Publishers, p. 102-119.

IgnaziP. (1992), “The Silent Counter-Revolution: Hypotheses on the Emergence of Extreme Right-Wing Parties in Europe”, European Journal of Political Research, 22(1), p. 3-34.

IgnaziP. (2003), Extreme Right Parties in Western Europe, Oxford, Oxford University Press.

InglehartR. (1977), The Silent Revolution: Changing Values and Political Styles among Western Publics, Princeton (nj), Princeton University Press.

Inglehart R. (1987), “Value Change in Industrial Societies”, American Political Science Review, 81(4), p. 1289-1303.

Inglehart R. (1990), Culture Shift in Advanced Industrial Society, Princeton, (nj), Princeton University Press.

Inglehart R. (1997), Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Countries, Princeton (nj), Princeton University Press.

Kalmijn M. (1994), “Assortative Mating by Cultural and Economic Occupational Status”, American Journal of Sociology, 100(2), p. 422-452.

KellyK. D. & Chambliss W. L. (1966), “Status Consistency and Political Attitudes”, American Sociological Review, 31(3), p. 375-382.

Lijphart A. (1968), The Politics of Accommodation: Pluralism and Democracy in the Netherlands, Berkeley (ca), University of California Press.

Lipset S. M. (1959), “Democracy and Working-Class Authoritarianism”, American Sociological Review, 24(4), p. 482-502.

LipsetS. M. (1981) Political Man: The Social Bases of Politics (Expanded Edition), Baltimore (md), Johns Hopkins University Press.

MiddendorpC. P. (1991), Ideology in Dutch Politics: The Democratic System Reconsidered (1970-1985), Assen, Van Gorcum.

MitchellR. E. (1966), “Class Linked Conflict between Two Dimensions of Liberalism Conservatism”, Social Problems, 13 (Spring), p. 418-427.

NieuwbeertaP. (1995), The Democratic Class Struggle in Twenty Countries, 1945-1990, Amsterdam, Thesis Publishers.

Nieuwbeerta P. (1996), “The Democratic Class Struggle in Postwar Societies: Class Voting in Twenty Countries, 1945-1990”, Acta Sociologica, 39(4), p. 345-384.

O’KaneJ. M. (1970), “Economic and Non-economic Liberalism, Upward Mobility Potential, and Catholic Working Class Youth”, Social Forces, 40(4), p. 499-506.

Olson D.V.A. & Carroll J. W. (1992), “Religiously-Based Politics: Religious Elites and the Public”, Social Forces, 70(3), p. 765-786.

RempelM. & ClarkT. N. (1997), “Post-Industrial Politics: A Framework for Interpreting Citizen Politics Since the 1960’s”, in Rempel M. & Clark T. N. (eds) (1998), Citizen Politics in Post-Industrial Societies, Boulder (co), Westview, p. 9-56.

ScheepersP.,Eisinga R. & Van SnippenburgL. (1992), “Working-Class Authoritarianism: Evaluation of a Research Tradition and an Empirical Test”, Netherlands’ Journal of Social Sciences, 28(2), p. 103-126.

Van der BrugW. & Fennema M. (2003), “Protest or Mainstream: How the European Anti-Immigrant Parties Have Developed into Two Separate Groups by 1999”, European Journal of Political Research, 42(1), p. 55-76.

Van der WaalJ., Achterberg P. & Houtman D. (2007), “Class Is Not Dead —It Has Been Buried Alive: Class Voting and Cultural Voting in Postwar Western Societies (1956-1990)”, Politics and Society,35(3), p. 403-426.

VeugelersJ. W. P. (2000), “Right-Wing Extremism in Contemporary France: A “Silent Counterrevolution”?”, Sociological Quarterly, 41(1), p. 19-40.

WeakliemD. (1991), “The Two Lefts? Occupation and Party Choice in France, Italy, and the Netherlands”, American Journal of Sociology, 96(6), p. 1327-1361.

Haut de page

Notes

1 . With the founding of ppr and psp the Netherlands witnessed the emergence of two new-leftist parties during the 1970’s. These two parties later on merged with the former Communist Party (cpn), which itself had changed unrecognizably due to the influence of the new spirit of the time, into the new party GroenLinks in 1989. GroenLinks has ever since remained the only new-leftist party of considerable size in the Netherlands.

2 . Examples of electorally successful new rightist-populist parties in Europe are the fpö in Austria, the Schweizerische Volkspartei (svp) in Switzerland, the Progress Party (frp) and the Danish People’s Party (df) in Denmark, the Progress Party (frp) in Norway, the Vlaams Blok (renamed to Vlaams Belang in 2004) in Flanders, Belgium, the Republikäner in Germany, Front National in France, and the Lijst Pim Fortuyn (lpf) and (later on) Geert Wilders’ Freedom Party (pvv) in the Netherlands.

3 . Nevertheless, Inglehart (1997, p. 4, 43, 47) holds that the shift towards “postmaterialism” is somehow “at the core” of a more general process of cultural change (see for skepticism about this: Houtman, 2003).

4 . Although the rightist-extremist Centre Party (later on: Centre Democrats) had managed to mobilize some electoral support in the Netherlands during the 1980’s, this was very limited indeed.

5 . We relied on either present occupation or last occupation (if retired or unemployed).

6 . As a log odds ratio is the natural logs of the odds ratio, these estimates can be conversed into odds-ratios quite easily. Negatively signed estimates are then transformed into odds ratios between zero and one, positively signed log odds ratios are transformed into odds-ratios between one and infinity. We chose to present the log-odds ratios in our tables as these are easier to interpret.

7 . In a previous publication we have already demonstrated that the same goes for voting for the new-leftist Greens (GroenLinks) rather than the old-leftist Socialist Party (sp) or vice versa (Achterberg and Houtman, 2006).

Haut de page

Table des illustrations

Titre Figure 1. Distinguishing class voting (path 2 x path 3) from cultural voting (path 4 x path 5).
URL http://journals.openedition.org/sociologie/docannexe/image/101/img-1.png
Fichier image/png, 10k
Haut de page

Pour citer cet article

Référence électronique

Dick Houtman et Peter Achterberg, « Two Lefts and Two Rights. Class Voting and Cultural Voting in the Netherlands, 2002 », Sociologie [En ligne], N°1, vol. 1 |  2010, mis en ligne le 26 avril 2010, consulté le 29 mars 2024. URL : http://journals.openedition.org/sociologie/101

Haut de page

Auteurs

Dick Houtman

Professor of Cultural Sociology at Erasmus University - Rotterdam
Houtman@fsw.eur.nl
Department of Sociology,
Faculty of Social Sciences, Erasmus University
P.O. Box 1738
3000 DR ROTTERDAM – Pays-Bas

Articles du même auteur

Peter Achterberg

Assistant professor of sociology at Erasmus University – Rotterdam
Department of Sociology,
Faculty of Social Sciences, Erasmus University
P.O. Box 1738
3000 DR ROTTERDAM – Pays-Bas

Articles du même auteur

Haut de page

Droits d’auteur

Le texte et les autres éléments (illustrations, fichiers annexes importés), sont « Tous droits réservés », sauf mention contraire.

Haut de page

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search