Long after it was proposed to be presented at IPSA 2014 World Congress it was approved for

Left-Right Ideology as a Dimension of Identification and as a Dimension of Competition André Freire Department of Political Science & Public Policies, ISCTE-IUL (Lisbon University Institute), Researcher CIES-IUL, Lisbon, PORTUGAL Emails: andre.freire@iscte.pt & andre.freire@meo.pt Paper to be delivered at the IPSA 23 rd World Congress of Political Science Panel on Determinants of party choice in a comparative perspective Chair: Professor Oddbjørn Knutsen, Department of Political Science, University of Oslo. Email: oddbjorn.knutsen@stv.uio.no Palais des congrès - 513d Thursday, July 24th - 11:00-12:45 Note: This article is forthcoming in The Journal of Political Ideologies: http://www.tandfonline.com/toc/cjpi20/current#.u7f_e7evdic Long after it was proposed to be presented at IPSA 2014 World Congress it was approved for publication in The Journal of Political Ideologies. Because it is still forthcoming, I upload only a short part of it (abstract, keywords, introduction and conclusions) at the IPSA website; interested readers should wait for the forthcoming publication in JPI. The full paper is only circulated among the members of the panel. 1

Abstract (155 words) There is some controversy about the dimensionality of the left-right (LR) distinction. Some authors argue for uni-dimensionality (LR structured by socio-economic issues). Other authors concede socio-economic issues are of fundamental importance in shaping the LR divide, but nevertheless argue for multi-dimensionality. Considering these controversies, we define our two major goals: to test the dimensionality of LR voting alignments among the mass public in 21 countries at the individual level, and to test the differentiation of the dimensionality of the LR divide as a dimension of identification and as a dimension of competition. Our main findings are as follows. First, as dimension of identification, LR voting has a multidimensional character: it is anchored in multiple sets of values orientations, both economic and non-economic. Second, the same cannot be said when we discuss it as a dimension of political competition : LR party-system polarization only has a significant and positive impact on left-right value voting for socio-economic values. Key phrases Left-Right Value voting Policy competition Party polarization Democratic consolidation Europe 2

Left-right ideology as a dimension of identification and of competition Introduction Ever since the French Revolution, the division between left and right has been of fundamental importance in mass politics, most of all in continental Europe, 1 although it is also recognized and used in parts of the world other than Europe and the United States. 2 This political divide has formed a categorization of ideologies, an instrument of classification of the political positions of the different parties, a code of communication (between political forces, mass media, and voters), and, finally, an instrument to direct voters in the interpretation of political phenomena and in decision making. 3 Additionally, in spite of all the theories about the end of ideology, the end of history, and about a certain overcoming of the division between left and right, the truth is that these same theories that argue in those directions have been clothed with an ideological character and, soon after being formulated/defended, were followed by the appearance of new ideological forms or by a renewed prominence of old ideologies. 4 Moreover, several studies have documented the importance of electors positions on the left-right scale as a defining factor of their voting choices, 5 even if some declining impact of left-right self-placement on the vote was recently found among younger Europeans after the fall of the Berlin Wall. 6 The importance of the left-right divide in modern mass politics notwithstanding, some controversies around the topic remain, namely about the uni-dimensionality or multi- 3

dimensionality of the concept in terms value and/or issue anchoring. For example, although he considers the liberty-authority divide to separate moderates and the extremists in each (left and right) camp, the political philosopher Norbert Bobbio argues for a major single criterion dividing the left and the right: the criterion most frequently used to distinguish between the left and the right is the attitude of real people in society to the ideal of equality. 7 In a similar vein, although from an empirical political science approach, Nöel and Thérien also argue for a uni-dimensional issue or value content associated with the left-right political distinction: these findings are very much in line with our understanding of the left and the right as anchored in distinct conceptions of equality in a democratic society, the first focused on unmet needs and solidarity, the other on opportunity and individual achievement. Each conception privileges specific social mechanisms: redistribution and state intervention for the left, competition and markets for the right. 8 Despite the primordial importance of the socio-economic issue/value divide to define the content of the left-right political distinction that most empirical studies document, empirical evidence has also shown that, especially among the mass publics, at least in some countries the issue content of the divide is multi-dimensional. Of course, the situation is variable across countries and time, particularly because of the diverse cleavage structures and their politicization in the different polities, but it has been shown that at the mass level, although the socio-economic content is usually the most important, the left-right divide correlates with religious and new politics issues, for example. 9 Moreover, research based on both expert surveys and party manifestos 4

show that while the socio-economic issues (socio-economic equality versus inequality, redistribution versus competition, the role of the state and the market in resource allocation) are usually the more important ones in structuring the left-right divide, the truth is that in many countries other non-economic issues also contribute to structure the political cleavage. 10 Considering all these controversies about the substantive content of the left-right divide both at the mass and at the party level the first objective of this article is to test the dimensionality of left-right voting alignments among the mass public of 21 European countries (in fact, 20 EU countries plus Israel). The second objective is to test whether there is any differentiation in the scale of the left-right divide as, first, a dimension of identification (that is, as an identity factor that structures citizens voting behaviour but that is by and large unresponsive to competing parties changing policy positions 11 ) and, second, as a dimension of competition (as a dividing line that structures party competition and that has or it is very likely to have some impact on the voting behaviour of citizens). 12 To fulfil the first objective we analyse the impact of several sets of issues and of their underlying value orientations upon citizens left-right voting behaviour in 21 countries. To fulfil the second objective we scrutinize some supply-side conditions, especially the level of party system polarization on left-right issues (the clarity of policy alternatives hypothesis ), that may account for variations in the impact of values on left-right voting across countries. 5

Although the impact of values on the vote is well documented, 13 the same cannot be said about the contextual conditions that may increase or decrease left-right value voting. To the best of our knowledge only one paper tests the impact of party system polarization on value voting. 14 However, this study has several limitations, particularly because it analyses only five established democracies. By testing the clarity of policy alternatives hypothesis in 21 countries, we can perform a much more extensive test of this theory. Moreover, although there are some studies that evaluate the contextual factors behind variation in the impact of left-right (LR) self-placement on the vote, 15 we also know that in a non-negligible number of countries left-right self-placement (LRSP) is only barely correlated with substantive policy preferences and with their underlying value orientations. 16 Thus, we know that in many countries LRSP cannot be taken as a good proxy for value orientations. Additionally, by using expert surveys measuring the position of parties in terms of LR issues, 17 we are able to test the impact of party polarization along the encompassing LR divide. 18 We also control for the age of the democratic regime to further specify the supply-side conditions on the impact of values on LR voters alignments. Here we want to contribute to clarifying the controversy among political philosophers and empirical political scientists about the dimensionality of the LR divide, which is here understood in terms of LR voter alignments in general. More specifically, as a further contribution to the debate, we want to evaluate the dimensionality of the LR political distinction both as a dimension of identification and as dimension of competition. The extension of the tests, which are made on scale never before 6

attempted (21 countries), and the simultaneous and combined use of mass survey data (from the European Social Survey 2002) and party system data (from the 2002 expert survey in Benoit and Laver), make this contribution to the debate of the dimensionality of the LR divide a rather robust and extensive one. 19 Our fundamental findings are the following. First: as dimension of identification, LR voting has a multi-dimensional character, i.e., it is anchored in multiple sets of values orientations, both economic ( socio-economic values ) and non-economic ( religion and social liberalism values, law and order values, immigration values, laissez-faire versus environmental protection values ). Second, the same cannot be said when we discuss it as a dimension of political competition : LR party-system polarization only has a significant and positive impact on left-right value voting for the case of socioeconomic values. The level of regime institutionalization also proved to be positively associated with LR value voting. Below we review the literature, following which we present the data and the methods. ( ) Concluding remarks In the literature on the dimensionality of the LR political distinction there is some controversy. Some authors argue for uni-dimensionality. For example, Norbert Bobbio argues that people s attitudes vis-à-vis equality and inequality are the single most important criteria in differentiating the left from the right; Nöel and Thérien argue that socio-economic issues are the fundamental dividing line differentiating the left from the right. 60 Other authors concede that socio-economic issues are of fundamental 7

importance in structuring the LR divide, but nevertheless argue for multidimensionality: that is, these latter authors consider there are other important sets of criteria (that vary across countries and across time), and that contribute significantly to differentiate the left from the right, although more so in some countries than in others. 61 Considering the above-mentioned controversies, we defined our two objectives and associated research questions here. First, we wanted to test the dimensionality of LR voting alignments among the mass public of 21 countries at the individual level. Second, we wanted to test the differentiation of the dimensionality of the LR divide as a dimension of identification (as a dividing line of voting behaviour) and as dimension of competition (as a dividing line structuring party competition that is expected to have a significant impact on voting behaviour). Accordingly, we formulated two central hypotheses (H1 and H2), and one auxiliary (H3). To test our hypotheses we used ESS 2002-2003 because the countries included in this comparative mass survey provide wide political and institutional variation, allowing us to test hypotheses on the role of contextual factors on the impact of values on LR voting alignments on a previously untried scale. This comparative survey also includes several indicators that allowed us to measure values associated with both old ( socio-economic values, religion and social liberalism values ) and new ( immigration values, law and order values, and religion and social liberalism values ) politics (we extracted five value factors from the ESS data using PCA), and this rich set of issue items allowed us to adequately test the dimensionality of the LR divide. Finally, the use of the ESS 2002-2003 allowed us to 8

match mass left-right voting with contextual information about LR party-policy competition furnished in the 2002 comparative expert surveys conducted by Benoit and Laver. 62 We mainly used HLM to explain individual and cross-country variations in terms of the impact of values on LR voting alignments. Hypothesis 1 (H1) was clearly confirmed because, as a dimension of identification, 63 LR voting is clearly an encompassing ideological divide that is anchored in multiple value orientations. To put it another way, even after controlling for social factors, LR voting alignments are determined not only by socio-economic value orientations but also by other non-economic value orientations (immigration, religion and social liberalism, law and order, and laissez-faire versus environmental protection). One fundamental implication of this finding is that, as dimension of identification structuring voting behaviour, LR is indeed an encompassing political device that is flexible enough to accommodate important political value conflicts in each country and/or in each epoch: for example, in countries where there are more immigrants and the topic is politicized, the values related with this subject are one important element to structure left-right voting; another example, in countries where religious and related lifestyle issues have significant importance in the political arena (for example, in Catholic, Orthodox or religiously mixed countries), the related value orientations have a significant impact on the left-right vote. The second hypothesis (H2) states that the greater the clarity of party-policy alternatives the more often citizens use value orientations to guide their LR voting 9

alignments. The third auxiliary hypothesis (H3) states that in newer democratic regimes the impact of values on LR voting will tend to be weaker than in the more consolidated democratic regimes, mainly due to much lower levels of party system institutionalization. Our major conclusions for the H2 and H3 are the following. As we said before, the LR vote significantly anchors multiple value orientations in a large majority of the 21 European countries examined. However, the same cannot be said about LR as a dimension of political competition. Here the LR divide seems to have a singledimensional character associated with socio-economic issues in the latter case we found that LR party-system polarization only had a significant and positive impact on mass left-right value voting in the case of the impact of socio-economic values on the vote (confirming H2). This latter conclusion is in line with another study that investigated a similar topic but with a different approach, even if the present study furnishes a much more extensive and robust test. Although using left-right placement on a 10-point scale as the dependent (and not left-right voting as we do here), analysing a much smaller number of countries (13 Western countries in 1990 and in 1999) than we do in the present paper, and using a less robust measure of party polarization (the LR distance between the two major parties, according to voters perceptions of party locations in each country) vis-à-vis ours (an index of polarization that takes into account the LR location of all the parties in the system and which is based on expert surveys), Freire also found that the impact of party polarization on LRSP value anchoring was only significant for socio-economic values. 64 10

The level of democratic consolidation and party system institutionalization also proved to be positively associated with the level of value voting in each country: the greater the level of democratic consolidation and party system institutionalization, the greater the level of LR value voting, thus confirming H3. In terms of the level of LR vote, this was confirmed in all situations except for the impact of law and order values and for religion and social liberalism values on the vote. Our general conclusion is that politics matters in explaining how important value orientations are to guide individual LR voting. The greater the clarity of party proposals, the more voters rely on values to guide them in their LR voting options, even if this relationship is also mediated by the type of values (it only worked as expected for socio-economic values ). Moreover, the institutionalization of the party system also proves important for voters to know where the parties stand and thus for them to be able to confidently rely on values to guide them in their LR voting options: although again this relationship is mediated by the type of values we are considering. Since value/issue voting is a precondition for the quality of political representation, all this means is that it is in the hands of the parties to increase the quality of the political linkage by presenting clear alternatives to the electorate. If the political elites really want the electorate s vote to be structured more by policy preferences and their underlying value orientations, they know what they need to do. They must present more differentiated policy alternatives as an efficient way of helping voters decide, 11

even if, according to our data on 20 EU countries plus Israel, they should focus on socioeconomic issue to be well succeeded on that objective. This would not only help voters choose (among policy options), it would also strengthen and consolidate the links between voters and parties, especially if the latter are responsible actors who, once in government, try to implement the campaign s policy proposals, thereby helping the parties (re-)build loyal electorates. In the case of new democracies, the important question for the future is whether the reduced level of LR value voting found here due to the lack of party system institutionalization will remain a permanent feature (even after stronger democratic consolidation), or if the new democracies will converge with the more consolidated ones. These conclusions were arrived at through the analysis of a very rich and diverse data set; however, in the future it will be desirable to increase the number of countries, particularly of those in new Europe, and to explore the impact of other social, political, and institutional characteristics upon value voting. 12

1 J. A. Laponce, Left and Right: The Topography of Political Perceptions (Toronto: University of Toronto Press, 1981). 2 R. J. Dalton, Social modernization and the end of ideology debate: Patterns of ideological polarization, Japanese Journal of Political Science, 7, 1 (2006), pp. 1-22; A. Noël and J.-P. Thérien, Left and Right in Global Politics (Cambridge: Cambridge University Press, 2008); A. Freire and K. Kivistik, - Western and non-western meaning of the left-right divide across four continents, Journal of Political Ideologies, 18, 2 (2013), pp. 171-199; A. Freire and K. Kivistik, Mapping and explaining the use of the left-right divide, Brazilian Political Science Review, 7, 3 (2013), pp. 61-89. 3 D. Fuchs and H.-D. Klingemann, The left-right schema, in M. K. Jennings, J. W. van Deth, S. H. Barnes, D. Fuchs, F. J. Heunks, R. F. Inglehart, M. Kaase, H.-D. Klingemann, J. J. A. Thomassen, Continuities in Political Action: A Longitudinal Study of Political Orientation in Three Western Democracies (Berlin: de Gruyter, 1990), p. 205. 4 A. Heywood, Political Ideologies: An Introduction (Basingstoke: Palgrave Macmillan, 2003), pp. 319-323; R. Eatwell, Conclusion: the end of ideology, in R. Eatwell and A. Wright, Contemporary Political Ideologies (London: Continuum, 2003), pp. 279-290. 5 R. Gunther and J. R. Montero, The anchors of partisanship: A comparative analysis of voting behaviour in four southern European countries, in N. Diamandouros and R. Gunther, eds, Parties, Politics, and Democracy in New Southern Europe (Baltimore, MD: The Johns Hopkins University Press, 2001), pp. 83-152; M. Franklin, T. Mackie and H. Valen, eds, Electoral Change: Responses to Evolving Social and Attitudinal Structures in Western Countries (Cambridge: Cambridge University Press, 1992). 6 W. van der Brug, Structural and ideological voting in age cohorts, in Z. Enyedi and K. Deegan-Krause, eds., The Structure of Political Competition in Western Europe (London: Routledge, 2011), pp. 172-193. 7 N. Bobbio, Left and Right: The Significance of a Political Distinction (Chicago, IL: Chicago University Press, 1997), pp. 72-79. 8 Nöel and Thérien, op. cit., Ref. 2, p. 44. 13

9 R. Inglehart, The changing structure of political cleavages in Western society, in R. J. Dalton, S. C. Flanagan and J. E. Alt, eds., Electoral Change in Advanced Industrial Democracies: Realignment or Dealignment? (Princeton, NJ: Princeton University Press, 1984), pp. 25-69; O. Knutsen, The partisan and the value-based components of left-right self-placement: A comparative study, International Political Science Review,18 (1997), pp. 191-225; Dalton, op. cit., Ref. 2; A. Freire, Party polarization and citizens left-right orientations, Party Politics,14, 2 (2008), pp. 189-209; Freire and Kivistik, op. cit., Ref. 2. 10 K. Benoit and M. Laver, Party Policy in Modern Democracies (London: Routledge, 2006); H.-D. Klingemann, A. Volkens, J. Bara, I. Budge and M. McDonald, Uniquely! The mapping of party policy movements in Central and Eastern Europe 1990-2003, in H.-D. Klingemann, A. Volkens, J. Bara, I. Budge and M. McDonald, eds, Mapping Policy Preferences II: Estimates for Parties, Electors and Governments in Eastern Europe, European Union and OECD 1990-2003 (Oxford: Oxford University Press, 2006), pp. 3-27. 11 H. Kitschelt, Z. Mansfeldova, R. Markowski and G. Toka, Post-Communist Party Systems: Continuity and Change (Cambridge: Cambridge University Press, 1999), p. 289. 12 The concept of dimension of identification and dimension of competition, which will be developed below, is taken from G. Sani and G. Sartori, Polarization, fragmentation and competition in Western democracies, in H. Daalder and P. Mair, eds, Western European Party Systems: Continuity and Change (London: Routledge, 1983), pp. 307-340. See also Kitschelt, Mansfeldova, Markowski and Toka, op. cit., Ref. 11, pp. 258-259 and 289. 13 S. M. Lipset and S. Rokkan, Cleavage structures, party systems, and voter alignments: An introduction, in S. M. Lipset and S. Rokkan, eds, Party Systems and Voter Alignments: Cross-National Perspectives (New York, NY: Free Press), pp. 1-64; S. Flanagan, Value change in industrial societies, American Political Science Review, 81, 4 (1987), pp. 1303-19; O. Knutsen and E. Scarbrough, Cleavage politics, in J. W. van Deth and E. Scarbrough, eds, The Impact of Values (Oxford: Oxford University Press, 1995), pp. 492-524; S. C. Flanagan and A.-R. Lee, The new politics, culture wars, and the authoritarian-libertarian value change in advanced industrial democracies, Comparative Political Studies, 36, 3 (2003), pp. 235-271. 14

14 O. Knutsen and S. Kumlin, Value orientations and party choice, in J. Thomassen, ed., The European Voter: A Comparative Study of Modern Democracies (Oxford: Oxford University Press, 2005), pp. 157ff. 15 C. van der Eijk, H. Schmitt and T. Binder, Left-right orientations and party choice, in Thomassen, ibid., pp. 167-191; R. Lachat, The impact of party polarization on ideological voting, Electoral Studies, 27, 4 (2008), pp. 687-698; B. Wessels and H. Schmitt, Meaningful choices, political supply, and institutional effectiveness, Electoral Studies, 28, 5 (2008), pp. 19-30; W. van der Brug, M. Franklin and G. Toka, One electorate or many? Differences in party preference formation between new and established European democracies, Electoral Studies 27, 4 (2008), pp. 589-600. 16 Knutsen, op. cit., Ref. 9; Freire, op. cit., Ref. 9. 17 Benoit and Laver, op. cit., Ref. 10. 18 By an encompassing LR divide we mean a political device, understood either as an instrument of citizens political orientation and/or as an instrument of political communication (between politicians, mass media and voters), that is flexible enough to accommodate within it different substantive meanings (the relevant political value conflicts in each country and/or in each epoch) across time and country. 19 European Social Survey, Round 1, 2002-2003, Dataset, ess.nsd.uib.no/2002; Benoit and Laver, op. cit., Ref. 10. 60 Bobbio, op. cit., Ref. 7; Nöel and Thérien, op. cit., Ref. 2. 61 Inglehart, op. cit., Ref. 9; Knutsen, op. cit., Ref. 9; Dalton, op. cit., Ref. 2; Freire, op. cit., Ref. 9; Freire and Kivistik, op. cit., Ref. 2; Benoit and Laver, op. cit., Ref. 10; Klingemann et al, op. cit., Ref. 10. 62 Benoit and Laver, ibid. 63 Sani and Sartori, op. cit., Ref. 11. 64 Freire, op. cit., Ref. 9. 15