You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Attitudes towards inequality in cross-national perspective : tracing the endurance of social class in… Haddon, Edward David 2019

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2020_may_haddon_edward.pdf [ 2.99MB ]
Metadata
JSON: 24-1.0386719.json
JSON-LD: 24-1.0386719-ld.json
RDF/XML (Pretty): 24-1.0386719-rdf.xml
RDF/JSON: 24-1.0386719-rdf.json
Turtle: 24-1.0386719-turtle.txt
N-Triples: 24-1.0386719-rdf-ntriples.txt
Original Record: 24-1.0386719-source.json
Full Text
24-1.0386719-fulltext.txt
Citation
24-1.0386719.ris

Full Text

  ATTITUDES TOWARDS INEQUALITY IN CROSS-NATIONAL PERSPECTIVE: TRACING THE ENDURANCE OF SOCIAL CLASS IN MODERN SOCIETIES  by Edward David Haddon M.A. (First Class Honours), The University of Auckland, 2012 B.A. (Hons), The University of Auckland, 2011 B.A., Simon Fraser University, 2009  A DISSERTATION SUMBITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSPOHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Sociology)  THE UNIVSERITY OF BRITISH COLUMBIA (Vancouver) December 2019  © Edward David Haddon, 2019  ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  Attitudes towards inequality in a cross-national perspective. Tracing the endurance of social class in modern societies.  submitted by Edward David Haddon  in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Sociology  Examining Committee: Sylvia Fuller, Sociology Supervisor  Rima Wilkes, Sociology Supervisory Committee Member  Tom Kemple, Sociology Supervisory Committee Member Gerry Veenstra, Sociology University Examiner Alan Jacobs, Political Science University Examiner   iii  Abstract This dissertation employs quantitative techniques including random and fixed effects linear regression to analyze a dataset comprised of 27 over 17 years to explore the extent to which people are critical of inequalities in modern societies. It investigates whether the positions people occupy in the social forces of production influence the extent to which they view inequality critically. This research also examines the interplay between placement within the class structure, political orientations and national-level factors in determining critical inequality views. The findings reveal that peoples’ disapproval towards inequality is strongly reflective of their class position. Yet, this also depends on their political persuasions and changes at different levels of income inequality. This is because the self-interests associated with class largely determines the views of those who are right-leaning, with working class conservatives significantly more likely to condemn inequality compared to their counterparts in upper class positions. The research also shows that the working classes are concerned with inequality in both unequal and more equal societies. As inequality increases, however, the views that the various classes have towards inequality begin to converge. Indeed, the results reveal that in contexts where inequality is high, upper class inequality views are more critical than the working class. This has possible policy implications, particularly as income inequality continues to grow.    iv  Lay Summary In this dissertation I employ quantitative techniques to analyze a dataset comprised of 27 countries over 17 years to explore the extent to which people are critical of inequalities in modern societies. I investigate whether people’s social class positions shape their views towards inequality. I also examine the interplay between these class positions, political views and national-level factors in determining their opinions towards inequality. I find that peoples’ disapproval towards inequality is strongly reflective of their class position. Yet, this also depends on their political views and changes at different levels of income inequality. I show that the working classes are concerned with inequality in both unequal and more equal societies. As inequality increases, however, the views that the various classes have towards inequality begin to come together. In fact, in contexts where inequality is high, upper class inequality views are more critical than the working class.     v  Preface Edward Haddon identified and designed the research program and analyzed all research data for this dissertation. Data for chapters 2 and 3 come from the International Social Survey Programme, Social Inequality Modules. Country level data for chapter 4 were obtained from the World Income Inequality Database, the Standardized World Income Inequality Database, and the United Nations Statistical Division Aggerate National Accounts. A version of chapter 2 has been published: Haddon, Edward. 2019. "How Class Shapes Critical Resentment toward Inequality: The Competing Forces of Stratification and Politics." International Journal of Sociology 49(4):241-63. A version of chapter 3 is currently under review in the Sociological Quarterly. Edward Haddon conducted all statistical analyses and wrote the entirety of both manuscripts.   Table of Contents Abstract ......................................................................................................................................... iii Lay Summary ............................................................................................................................... iv Preface ............................................................................................................................................ v Table of Contents ......................................................................................................................... vi List of Tables ................................................................................................................................ ix List of Figures ................................................................................................................................ x List of Abbreviations ................................................................................................................... xi Acknowledgements ..................................................................................................................... xii Dedication ................................................................................................................................... xiii Chapter 1: Introduction ......................................................................................................... 1 1.1. Social Class: What it is and what it is not ........................................................................ 3 1.1.1. Similarities, Constraints and Convergence ............................................................... 5 1.2. Death of Class or Class Endurance? ................................................................................ 6 1.2.1. Subjective Implications of Class ............................................................................... 8 1.3. Education and Class-based Divisions in Viewing Inequality .......................................... 9 1.4. Focusing in on Critical Views Towards Inequality: Rationale and Justifications ......... 13 1.5. Outline of the Dissertation ............................................................................................. 16 1.5.1. Connecting the Empirical Chapters ........................................................................ 16 1.5.2. Summary of the Findings ........................................................................................ 16 Chapter 2: How Class Shapes Critical Views Towards Inequality: The Competing Forces of Stratification and Politics .......................................................................................... 19 2.1. Introduction .................................................................................................................... 19 2.2. Theorizing Class, Education, and Attitudes Towards Inequality ................................... 22 2.3. Class and Opposition to Inequality ................................................................................ 25 2.4. The Interplay between Class, Education and Attitudes Towards Inequality ................. 27 2.5. Class and Party ............................................................................................................... 28 2.6. Methods .......................................................................................................................... 30 2.6.1. Data ......................................................................................................................... 30 2.6.2. Predictors ................................................................................................................ 31 2.6.3. Outcome .................................................................................................................. 34 2.6.4. Additional Controls ................................................................................................. 36 2.7. Analytical Strategy ......................................................................................................... 38 2.8. Results ............................................................................................................................ 39 vii  2.9. Discussion and Conclusion ............................................................................................ 48 Chapter 3: Class, Partisanship and The Great Recession: Conflicting Influences on Attitudes Towards Inequality During Economic Crises ......................................................... 53 3.1 Introduction .................................................................................................................... 53 3.2 Public Opinion Shifts to Economic Shocks ................................................................... 56 3.3 Opposition and Contentment to Economic Instability ................................................... 58 3.4 Theorizing Class Responses to Economic Crises .......................................................... 60 3.5 Politics During Crisis ..................................................................................................... 62 3.6 Class Politics During Crises ........................................................................................... 63 3.7 Methods .......................................................................................................................... 64 3.7.1. Data ......................................................................................................................... 64 3.7.2. Dependent Variable ................................................................................................ 65 3.7.3. Key Independent Variables ..................................................................................... 65 3.8 Analytical Strategy ......................................................................................................... 73 3.9 Results ............................................................................................................................ 74 3.10 Discussion ...................................................................................................................... 88 3.10.1. Class Responses to Economic Crises .................................................................. 88 3.10.2. Economic Shocks and Class Politics ................................................................... 90 Chapter 4: Class and Country: The Interplay Between Social Class and Societal-level Factors in Determining Citizens’ Attitudes Towards Inequality ........................................... 92 4.1. Introduction .................................................................................................................... 92 4.2. Class and Country .......................................................................................................... 94 4.2.1. Economic Dispersion and Views Towards Inequality ............................................ 94 4.2.2. Economic Development and Views Towards Inequality........................................ 97 4.2.3. Theorizing the Interplay between Class, Objective Inequality and Economic Growth on Attitudes Towards Inequality .............................................................................. 99 4.3. The Class Politics of Welfare Regimes ........................................................................ 101 4.3.1. The Spectre of Communism and Views Towards Inequality ............................... 104 4.4. Data .............................................................................................................................. 105 4.4.1. Dependent Variable .............................................................................................. 109 4.4.2. Individual-level Variables ..................................................................................... 109 4.4.3. National-level Variables ....................................................................................... 109 4.4.3.1. Income Inequality ..................................................................................................... 109 4.4.3.2. Economic Development ............................................................................................ 110 4.4.3.3. Regime Type ............................................................................................................. 111 viii  4.5. Analytical Strategy ....................................................................................................... 111 4.6. Results .......................................................................................................................... 113 4.7. Discussion and Conclusion .......................................................................................... 127 Chapter 5: Conclusion ........................................................................................................ 131 5.1. Re-animating Class ...................................................................................................... 131 5.2. Research Significance and Implications for Class Politics .......................................... 132 5.3. Future Research ............................................................................................................ 134 5.4. Class as a Psychological Phenomenon ......................................................................... 136 Bibliography .............................................................................................................................. 139 Appendices ................................................................................................................................. 151 Appendix A ............................................................................................................................. 151 Appendix B ............................................................................................................................. 172 Appendix C ............................................................................................................................. 180   ix  List of Tables Table 2.1 Descriptive statistics for each country ........................................................................................ 37 Table 2.2. Fixed effects of linear regression models and interactions between class, education and political affiliation in critically inequality views ........................................................................... 43 Table 2.3. Average marginal effects for Model 4 ....................................................................................... 46 Table 2.4. Average marginal effects for Model 6 ....................................................................................... 46 Table 3.1. Collapsing the eight-class model into a three-class version....................................................... 67 Table 3.2 Class compositions for each country and year (proportions) ...................................................... 68 Table 3.3 Proportions for key variables ...................................................................................................... 71 Table 3.4 Results from linear fixed effects regression models and interactions between year, class and political affiliation in inequality views .......................................................................................... 81 Table 3.5 Average marginal effects for class relationships to attitudes towards inequality across years (Model 3) ....................................................................................................................................... 84 Table 3.6 Average marginal effects for party affiliation’s relationship to attitudes towards inequality across years (Model 4) ................................................................................................................... 84 Table 3.7 Average marginal effects for class and party affiliation across years (Model 5) ........................ 84 Table 4.1 Descriptive statistics for each country and year ........................................................................ 107 Table 4.2 Mixed effects linear models predicting critical views towards inequality (Standard errors in parentheses) ................................................................................................................................. 119 Table 4.3 Average marginal effects of the classes at various values of Gini. Results are from Model 2 . 123 Table 4.4 Average marginal effects for classes across regime types. Results are from Model 4 ............. 123 Table 4.5 Average marginal effects for class across levels of Gini. Results are from Model 5 ............... 124 Table 4.6 Average marginal effects for class across regime type. Results are from Model 5 .................. 124    x  List of Figures Figure 2.1 Class compositions for each country (proportions) ................................................................... 33 Figure 2.2. Linear predictions of viewing inequality critically by class and party affiliation .................... 47 Figure 3.1 Mean inequality views by survey year ...................................................................................... 75 Figure 3.2 Mean inequality views by social class over time ....................................................................... 76 Figure 3.3 Linear predictions of inequality views by class (1992, 1999, 2009) ......................................... 85 Figure 3.4 Linear predictions of inequality views by party affiliation (1992, 1999, 2009) ........................ 86 Figure 3.5 Linear predictions of inequality views by class and party affiliation (1992, 1999, 2009) ........ 87 Figure 4.1 Country averages on subjective inequality index by income inequality (Gini) ......................... 96 Figure 4.2 Country averages on subjective inequality index by economic growth (GDP) ......................... 98 Figure 4.3 Critical views towards inequality by Gini. Predictions are from Model 2 .............................. 125 Figure 4.4 Critical perceptions of inequality according to regime type. Predictions are from Model 4 ... 126     xi  List of Abbreviations ISSP  International Social Survey Programme SWIID  Standardized World Income Inequality Database TGR  The Great Recession  xii  Acknowledgements  I would like to acknowledge the faculty and student body in the Department of Sociology at the University of British Columbia for the creation of a productive and stimulating environment. I owe particular thanks and would like to express my great appreciation to Sylvia Fuller for her valuable and constructive suggestions during the development, analysis and writing stages of this research. Her encouragement and useful critique have led me to produce a thesis that I can be proud of. Tom Kemple provided immense guidance into the theoretical underpinnings of this research. I would also like to thank Rima Wilkes for her willingness to read and provide feedback on various chapters of the thesis. My appreciation should also be extended to GESIS – Leibniz Institute for the Social Sciences in Germany—for providing access to the ISSP dataset and hosting me as a visiting researcher in 2015. Finally, but most certainly not least, I would like to acknowledge my partner Olya, whose unconditional support has in large part made the completion of this thesis possible. Your devotion and patience provided a sense of stability that I cannot fully express here. Thanks to my two little whippersnappers, Catherine and Aneira, for putting a smile on my face every time I emerged from my office in the bowels of our home. I am also grateful to the rest of my family for their moral support and encouragement over the past several years.  xiii  Dedication I would like to dedicate this dissertation to the memory of my grandfather, Dr. John Frost. Your wisdom and creativity are qualities I can only hope to aspire to.   1  Chapter 1: Introduction This dissertation is intended to contribute a class analysis of how people view inequality in their society. It is informed by Marxist and Weberian theory which I employ to generate a measurement of social class situated firmly in the social relations of production. Within sociology, social class has traditionally been viewed as the discipline’s first key discovery. Indeed, for some time, the role that social class plays in shaping perceptions of social phenomenon was generally accepted and undisputed (Kelley and Evans 1995). However, scholars in the past few decades have questioned whether social class remains a relevant concept for understanding Western societies (Beck and Beck-Gernsheim 2002; Kingston 2000; Pakulski and Waters 1996). Critics of class argue that there are other, more important sources of interest which impact individual views about the world and that class is no longer useful in understanding contemporary social phenomenon. This argument is regularly offered by those who suggest that class has lost its political significance. My PhD dissertation is provoked by these claims.  A common argument exists that social class has little impact on economic values, such as attitudes towards inequality (Bauman 1982; Beck and Beck-Gernsheim 2002; Clark and Lipset 1991; Inglehart and Welzel 2005; Inglehart and Norris 2017; Nisbet 1959). Recent developments on the political sphere challenge this idea. There is reason to believe that, regardless of whether class has declined in importance, it is certainly becoming relevant again. In 2008, the world faced a global financial crisis, which left millions of people unemployed and many struggling with economic insecurity and reductions in incomes. Average incomes stagnated or fell in most countries post-crisis, from the years 2007 through to 2011/2012 (OECD 2014). In response to these events, the global community witnessed a surge in public backlash against the capitalist system which culminated in the Occupy Wall Street movement; a movement which pushed the issue of growing inequality into mainstream discussion and highlighted serious issues with the capitalist system itself.  Classic Marxist theory suggests that economic change can create conditions for class resurgence. Marx noted that there is a tendency of capitalism towards crisis and class antagonism. Indeed, Piketty 2  (2014) more recently has demonstrated the inclination for capitalism to perpetuate inequality rather than reduce it. Whether or not an economic slump increases the importance of class with respect to attitudes towards inequality depends on a whole set of complex factors, including the nature of the crisis and the constellation of societal forces contributing to it. Nevertheless, among the social groups that individuals are located in, arguably the most relevant for responding to economic recessions is social class. This is because class locations structure an individual’s economic vulnerability, with different social classes experiencing different likelihoods of negative effects during periods of economic uncertainty.  While concerns with rising inequality have occurred in some countries, others may not have experienced similar trends, which makes it necessary to explore perceptions of inequality cross-nationally. Moreover, it is unclear whether all people within a society will perceive inequality similarly. Did we move towards more conflict laden societies as the recent protests surrounding inequalities would suggest? Does income inequality have the same polarizing effect on perceptions of inequality across groups within countries? How and why might the relationship between class and perceptions of inequality itself vary across countries? This doctoral research will address these questions using three waves of cross-national data on social inequality. First, I consider how social class may account for differences in views towards inequality across countries. Second, I trace aggregate trends in perceptions of inequality over time in various countries; exploring how these views were impacted by the global economic crisis in 2008. Finally, I assess the impact of macro-level characteristics including income inequality, economic growth and regime type on class perceptions of inequality. I show that many people are indeed concerned with inequality in their society and this anxiety is shaped by social class. I provide insight on the relationship between class positioning and attitudes towards economic inequality and how this relationship is moderated by additional factors at the individual and societal levels. I contribute evidence that concern with inequality is strong in most societies but also patterned by social class and reveal how this pattern changes under certain economic and political 3  circumstances. Not only does class play a role in shaping perceptions of inequality in various countries, it varies greatly depending on political persuasions.  1.1. Social Class: What it is and what it is not  Class is a long-standing fascination in sociology which has its origins primarily within Marx’s writings. Marx ([1867] 2012: 190), however, never systemically provided an answer to the question – ‘What constitutes a class?’ – he posed near the end of the final chapter of Capital Volume 3. Marx did not produce a “class schema” for the capitalist mode of production; instead positing a binary (bourgeoisie and proletariat) that operates as a dialectic. His primary concern was the identification and political construction of class interests rather than the identification of individual class position.  Marx’s classes designate the objective positions in the social division of labour. He distinguished one class from another on the basis of two criteria, (1) ownership of the means of production and (2) the purchase of the labour power of others. He points out that there also exist a number of intermediate classes which comprise the class structure which later Marxists have labelled “contradictory positions” (Wright 1997). How one earns a living (sources of revenue) and the degree to which one has control of the means of production are also crucial in determining one’s class position (Marx [1867] 2012). Essentially, Marx’s classes are premised on three elements: (1) classes are comprised of common positions; (2) these positions are understood relationally; (3) and the relationships between the positions are situated in the mode of production. Viewed in this way, later Marxists have established theoretical classes in terms of common structural positions situated within the socio-relational organizations of production (Wright and Perrone 1977). In essence, the forces generating Marxist classes constitute the positions within social relations of production and must also always be understood relationally; that is by their relationship to other classes. As the capitalist system evolves, these positions become increasingly delineated between the two polar opposites of the bourgeoisie and the proletariat. The exploitive relations of production are the primary 4  generator of class divisions, yet a strictly two-class Marxist model is not conducive to empirical analysis as there are so few “true” capitalists in a Marxist sense. Certainly, this empirical deficit is often why Marxists can be accused of a form of determinism, which was most notably critiqued by Weber. Weber’s contribution to class shifts the argument from the binary created by Marx to a more multi-dimensional argument relating to life-chances and markets. He was concerned with what was taking place in the market place or in exchange relations rather than the forces which generated the systems of production (Allen 2004). Weber defines classes as constituting people who have life chances in common and are determined by their power to dispose of goods and skills for the sake of income. For Weber ([1914] 2012: 312), the crucial aspect of class situation is the market, ‘the factor that creates “class” is unambiguously economic interest, and indeed, those interests involved in the existence of the market’. An individual’s class situation is determined by their ‘chance in the market’; ultimately one’s ‘class situation’ is based upon ‘market situation’ (Weber [1914] 2012: 312). Rather than understanding classes as social, Weber interprets them as ideal-typical concepts which are non-social groupings of individuals sharing similar economic or exchange relations.  A primary difference between Marx and Weber’s understanding of class is that Weber interprets production from the standpoint of market exchanges where various assets are exchanged; whereas Marx sees production in terms of the level of exploitation it generates. Thus, an important feature distinguishing Marx’s conception of class from Weber’s is the former’s linkage to the problem of class emancipation and a general theory of historical trajectory; while the latter is more explanatory and emphatically non-political. In this way, Weber’s understanding of class contains the remnants of a Marxists relational model, yet it is decidedly less deterministic.  In addition to promoting an empirically measurable relational notion of class, Weber’s writings are useful in so far as he draws a distinction between class and status situation. Weber understood classes to be situated within the economic order and status groups in the social order. The former is economically determined, whereas the latter is determined by a ‘social estimation of honour’ (Weber [1914] 2012: 314). 5  Weber ([1922] 1978) argued that during periods of relative market stability society tends to be stratified by status. However, as crises ensue or during periods or economic transformation “class situation” is pushed to the foreground as stratification based upon status recedes. In this way, a society stratified by either class or status can prevail depending on the particular form of societal division.  1.1.1. Similarities, Constraints and Convergence  Often Marxist and Weberian theorisations of class are regarded as inherently oppositional, epitomised in the sustained oeuvres of Goldthorpe (1992; 1996; 2002; 2006) and Wright (1993; 1996; 1997; 2000; 2002). This approach has its benefits, but it needs to be contextualized and the differences between the two approaches to class and the connection to individual subjectivities need to be balanced with the similarities. Both Weber and Marx share a view that capitalism creates the conditions needed for class associations and class struggles to occur. Being in possession of the means of production or one’s own labour power may shape individual views because possessing these contours the various alternatives individuals face when seeking material well-being. Wright (2002) for instance, reveals that both Marx and Weber saw that there was a strong propensity for the material interests of particular class positions to shape the behaviour of the individuals occupying those positions.  The debate between whether or not to operationalize class in terms of a Marxist or Weberian relational measure in empirical research is now largely irrelevant. Indeed, Wright (1997: 37) states that his schema ‘does not dramatically differ from the class typology used by Goldthorpe’. Nonetheless, when incorporating a relational measure of class, one is still faced with a decision between opting for a theoretically or empirically strong approach. Comparatively, it is felt here, where Marx and later Wright excel, is in precisely their commitment to something larger – in a theory that posits that only by understanding capitalist society can we hope to liberate ourselves from its grip. Weberian approaches may offer greater empirical possibilities, however they fall short in terms of the emancipatory potential offered by Marx (Erikson and Goldthorpe 1992). In this way, it is less a choice between theories of society, but between a theory and nontheory (Wright 1998: 320). A broader macro focus is, for the most part, absent 6  from Goldthorpe’s (2008: 350) Weberian approach to class, as a ‘narrowly defined concept of class’ is preferred.  A contracted view of class, however, restricts the focus to only one type of mechanism, reducing ‘the critical content of class analysis’ and ultimately blunting ‘its moral and political relevance’ (Wright 2008: 329). Therefore, the approach adapted here consists of a Weberian inspired operationalisation of class for the analysis of variations in individual views towards inequality, while retaining the more abstract Marxist concept of class to theorize the important role that social class plays in shaping these views. This route permits a Weberian assessment of class position which satisfies a Marxist theory in which class is understood to generate subjectively salient groups. 1.2. Death of Class or Class Endurance?  Some post-modern theorists assert that a newer variant of the social structure has emerged not foreseen by Marx or Weber. The new ways of living under post-modernity have apparently revealed ‘dynamic possibilities for a reorganization of social relations, which cannot be adequately comprehended by following either Marx or Weber’ (Beck and Beck-Gernsheim 2002: 36). Near the end of his career, along with Terry Clark, Seymour Martin Lipset (1991) began to question the importance of class in understanding individual perceptions and attitudes. An argument persists that while the structural importance of class may endure, the importance of class in terms of affecting people’s perceptions has weakened. Social class is argued to have lost its political significance as the class structure has becomes less distinct. Some scholars suggest that individualism and modernisation has meant that social class is of little importance to politics and subjectivities of modern society (Beck and Beck-Gernsheim 2002; Inglehart 1990; Kingston 2000; Pakulski and Waters 1996). It is, in part, the apparent reduction in class voting and class identities, which has led some scholars to regard class as an insignificant feature of identity and society. 7  Beck (2002: 203) takes his critique of class further by suggesting that it is a ‘zombie category’, in that the concept is still employed and alive within academia but in reality it is dead and outdated. He suggests that the dynamics of the labour market, bolstered by the welfare state, have rendered class unimportant contemporarily. Beck contends that a newer variant of the social structure will emerge not foreseen by Marx or Weber in which class society will dwindle in significance as an individualized society of employees reigns supreme. He goes on to argue that it ‘is very difficult to work in a rich empirical way with class categories . . . If you are interested in what is going on in people’s minds and the ways of life they are leading, you have to get away from the old categories’ (Beck and Beck-Gernsheim 2002: 207). However, Beck (2002: 34) advances his obituary by employing a pseudo-Weberian understanding of class, which he claims is comprised of different elements including ‘material conditions dependent on specific market opportunities, the effectiveness of tradition and of pre-capitalist lifestyles, the consciousness of communal bonds and of barriers to mobility, as well as networks of contact’ all wrapped up in the ‘shared life experiences mediated by the market and shaped by status’. This, as will be elaborated, is a weak conception of a Weberian understanding of class but one that is all too common in research claiming to explore the effects of class on a whole host of attitudes. One is then left wondering how Beck can claim that class is a walking corpse, entirely opaque to most individuals as a source of identity, when, as Atkinson (2007: 358) suggests, ‘he is not sure what the living, breathing body looked like’. Beck is regrettably not alone in proclaiming that class is obsolete because people no longer seem to be experiencing it as a salient social identity or are ambivalent towards it. Others have declared that there are other more important sources of interests which affect subjectivities (Bauman 1982; Clark and Lipset 1991; Kingston 2000; Pakulski and Waters 1996). Theorists in this mind-set regard the apparent weakness of class as indicative of a period of major social and cultural change that is decidedly different from the classed societies of the past (Inglehart and Flanagan 1987; Inglehart 1990). Class divisions are 8  said to be less important foundationally in terms of social identities, as other characteristics such as race, ethnicity, gender and sexuality are increasingly significant.  1.2.1. Subjective Implications of Class  Marx felt that capitalism would generate increasing polarization amongst those who own the means of production at one end and those who must sell their labour to survive at the other. ‘Society’, he wrote with Engels, ‘is more and more splitting up into two great hostile camps, into two great classes directly facing each other (Marx [1848] 2012: 157). Alongside this, class antagonisms would become increasingly simplified as membership in a particular class included the embodiments of the particular classes and the associated class interests. The increasing gulf between the upper and lower classes continues to expand; the popular mantra of the 99% versus the 1% speaks to this issue. Mass reaction to the “global economic crisis” illumines a degree of heightened awareness to class-based issues. The cultural setting appears to be ripe for an emergence of social action on the basis of class situation which both Marx and Weber suggested was possible.  Those suggesting that class is “dead” mistakenly purport that lack of identification implies the concept’s demise, and also dismiss empirical research revealing that class continues to generate subjectively salient identities. Recent analyses have supported the claim that class does impinge on individual subjectivities, with scholars finding relationships between class, life chances and political attitudes (Kelly and Enns 2010; Koçer and van de Werfhorst 2012; McCall and Manza 2011). Indeed, the failure of class to predict value preferences is not in opposition to a Weberian understanding of class as ‘[t]he emergence of an association or even of mere social action from a common class situation is by no means a universal phenomenon’ (Weber cited in Gane 2012: 102). Weber’s “economic classes” are forged within markets, which are sites of competition and are not necessarily comprised of a sense of belonging; feelings of solidarity and closeness with others in similar situations may not follow. There are then various class situations which could arise which means that it is unlikely that a common consensus or consciousness will follow: ‘people might experience a similar economic situation, [yet] this 9  commonality in itself does not necessitate the emergence of a meaningful social relationship’ (Gane 2012: 102). The ideal environment for the spread of class consciousness within groups occurs (1) during conflict between ‘immediate economic opponents’; (2) when there is a large group of individuals in the same class positions; (3) when the ability to mobilize individuals in similar class positions is simple or unrestrained; and (4) if class groups are directed towards particular aims which they are knowledgeable of but that are identified by individuals outside their class (Weber [1922] 1978: 305).   By suggesting that political consensus or social solidarity does not necessarily follow from one’s class situation, Weber allows us to bypass those who claim that disidentification implies that class, as a concept, is dead. While we may be able to theoretically counter anti-class scholars, we must also be careful not to claim that class position always generates subjectively salient groups. Nevertheless, capitalism is lifting the veil on the class structure of society and worker associations and shared sentiments based on these are becoming increasingly likely. It is useful at this point to return to Marx. 1.3. Education and Class-based Divisions in Viewing Inequality “the most rapid possible growth of capital . . . does not remove the antagonisms between his interests and the interests of the bourgeoisie . . . If capital is growing rapidly, wages may rise; the profit of capital rises incomparably more rapidly. The material position of the worker has improved, but at the cost of his social position . . .  the forest of uplifted arms demanding work becomes ever thicker, while the arms themselves become ever thinner” (Marx [1847] 2012:187-89). Marx felt that there is a tendency for society to be divided as inequality increases. The working class grows ever larger and people begin to compete amongst themselves for any slice of the dwindling wage pie: antagonisms within and across classes increase. Is class truly in decline in terms of people’s perceptions of this vast gulf of inequality as death of class theorists assert?  10  I adhere to the Marxist premise that control over the material interests rooted in the social forces of production connected to the class structure exist and that these exist independently of the actual inhabitants of the class locations within this structure. Assuming this to be the case, I contend that there is the possibility for the individuals residing in these locations to develop perceptions of society that are relatively consistent with the interests of the class that is objectively theirs. This possibility assumes that all members of a particular class location may not share the same perceptions; but that there is the probability that they will share perceptions which are consistent with the objective interests connected to their class locations, and that the probabilities are higher than for the members of other class locations. While individual attributes may affect the strength of the association between the class structure and viewing inequality, this association will be primarily based on the properties of the class structure.  Proponents of class have attempted to explain the discrepancy between class and individual subjectivities by suggesting that rising prosperity has ushered in a period of a decline in social class shaping value preferences (Inglehart and Rabier 1986; Inglehart and Flanagan 1987; Inglehart and Abramson 1994), that cultural aspects mediate the construction of class boundaries (Bottero 2010; Lamont 2000), or that class voting has not declined, rather that political parties of the left and right have converged more to the centre in terms of their political ideology (Andersen, Yang and Heath 2006; Evans and Tilley 2012b). These explanations do provide insights into why economic development, party affiliation and objective class might vary in the extent to which they may influence perceptions of society. However, they have limitations with respect to understanding how each can operate in shaping views towards inequality across time and place. While Marxist and Weberian understandings of the degree to which class might endure differ, they ultimately speak to the importance of key underlying questions: when is class most salient? Why might it “matter” more in some times and places than others?  During periods of relative market stability, for instance, society tends to be differentiated by status, which Weber ([1922] 1978:38) suggested was connected to the “social estimation of honor” attached to certain positions; while class, which is grounded in the social relations of economic life, is 11  placed on the backburner. During periods of economic turmoil, the situation is reversed, with class pushed to the fore as stratification based upon status recedes. From a Weberian perspective, it is plausible to propose that periods of economic instability represent an instance where the class structure of society may be growing in importance, while the status structure recedes.  The argument and analyses in this dissertation originate from a theoretical perspective that societies are characterized by two distinct forms of stratification. This implies, that in order to fully understand the effects of stratification on public opinions towards inequality, it is necessary to differentiate between the two forms. Weber’s emphasize on the social relations situated within labour markets and production units are central features of the class system. Education is regarded as a core aspect of the status system. For Weber, status is a term used to describe how positive and negative privileges associated with various elements such as race, gender, and education work. Status is a claim to social esteem based on 1) manner of living ones life; 2) formal training and socialization; and 3) prestige of ancestors and profession (Weber 1922/1956/1964 cited in Waters and Waters 2016). Status refers to structures where the interests is in resources related to honour, cultural distinctiveness and shared world views (Waters and Waters 2016). Education, for Weber, enabled the student to cultivate status. In doing so, Weber links the distinction between intellectual inquiry and individual self-formation. Education is an achieved status and the cognitive skills that higher levels generate are embodied within the individual. The cognitive skills needed to be aware of the existence of income differences and the ability to see the social world more broadly are inculcated within individuals through their achieved level of education. In this way, education shapes values about inequality and it also provides individuals knowledge of it. Education occupies an ambiguous position in the role that it plays in shaping inequality views. Broadly speaking, there are two theories on the mechanisms through which education impacts economic attitudes. The first is a more cognitive model, which maintains that education increases cognitive skills and broadens a persons horizons by providing them the ability to perceive social issues from different 12  points of view (Hyman and Wright 1979). Increased education embodies a source of access to practical knowledge of the social world. The highly educated possess the capacity to “think clearly”, provided by the unique talents and garnered skills acquired though academia. Those who are better informed also may provide different explanations for economic inequality, possibly stressing various social causes (e.g. inequalities in educational opportunities, governmental policies, discrimination and so on) more so than those or are less highly educated (Bartels 2005).  The second, but related mechanism, pertains more to socialization, in that education socializes students in their various political orientations. Within any status group, Weber felt that certain ideologies and belief systems emerge. In this view, an advanced education is thought to embody within the possessor a more enlightened perspective. In higher education, the take-for-granted assumptions of the world are often challenged, and instead, social reality is a product of human actions (Gabennesch 1972). For instance, in the American context, McCall (2013:110) finds that those with more education are more likely to desire less inequality. She suggests that education “appears to reflect a socially liberal orientation toward inequality or access to specialized information” (McCall 2013:110).  Others hold a more pessimistic view of how status socializes individuals. For Ridgeway (2014:3), status is “based on widely shared beliefs about the social categories or “types” of people that are ranked by society as more esteemed and respected compared to others”. In other words, status often imbues public recognition of social worth. These shared beliefs shape individual expectations of themselves and their actions in social contexts. Those who are highly educated, for instance, may hold status beliefs containing implicit assumptions that they are more competent, more deserving of money or promotions, than those with lower levels of education. In this view, education may instead provide one with intellectual training and a more sophisticated grasp of information, enabling recipients to first perceive inequality in their society and then defend their interests more forcefully (Jackman and Muha 1984). As a result, those who are highly educated may recognize that income differences are greater, yet they may not perceive them as too large. Indeed, Bartels (2005) finds that “better-informed people” have more 13  pessimistic views about the nature and consequences of inequality in America; while Koçer and van de Werfhorst (2012) find that higher education is associated with more favourable perceptions of the current income distribution.  1.4. Focusing in on Critical Views Towards Inequality: Rationale and Justifications  The goal of the dissertation is to explore the relationship between economic conditions at the individual and country level and attitudes towards inequality. These economic attitudes are a key aspect which, one would assume, should be impacted by social class. Scholars who contend that class is dead, however, often do so by suggesting that the association between class and value preferences towards economic issues has declined (Beck and Beck-Gernsheim 2002; Inglehart 1990; Pakulski and Waters [1996] 2008). This post-materialist thesis suggests that the rising prosperity that has accompanied modernisation and economic development has meant that factors based on the social relations of production play a reduced role in shaping attitudes towards inequality, ushering in post-materialist values. The views people hold towards inequality therefore represents an important area with which to explore the continued salience of class. Perceptions of inequality, also referred to as subjective inequality, represents a multidimensional phenomenon which not only include perceptions of inequality, but also beliefs about what constitutes fair inequality and judgements about existing inequality (Janmaat 2013). People often live with a tension between how they perceive the world and what they think the world should be, thus they often adjust their perceptions to their normative beliefs (Gijsberts 2002; Sachweh and Olafsdottir 2012). For example, those who agree with the notion that merit should determine income are likely to accept larger income differences compared to those who maintain that income should be based on equality. Individuals who maintain that current inequalities are fair may also hold negative views towards disadvantaged groups due to their conviction that these marginalized groups have not “tried hard enough” or “get what they deserve”.  14  Exploring perceptions of inequality is important because a condition for achieving the legitimacy and consensus of a particular economic system is having a certain degree of consensus regarding the extent of material inequality within a particular society (Wilkinson and Pickett 2018). A deviation between perceptions of desirable levels and the actual level of economic inequality may indicate a crisis of legitimation in the political arena. Public opinions towards income inequality tells us about the extent to which current income distributions in a society are regarded as legitimate by the populace and may determine normative judgements about what the distribution should be (Saar 2008). For example, preferences for redistributive policies largely rest on whether individuals think current levels of inequality are legitimate (Kenworthy and McCall 2008). Scholars show that those who are more concerned about income differences are also more likely to support redistributive policies (McCall and Kenworthy 2009). The survey item employed in this dissertation to measure attitudes towards inequality does not directly measure perceptions of how much inequality there is. What it does tap into is the awareness of the degree of income inequality and the associated degree of fairness about the perceived level of income inequality. It is a complex question because it combines two aspects: the perceptions of inequality and the opinions towards fairness of inequality. People need to see inequality in their society first and also hold critical inequality views in order to justify a possible state response. In this way, inequality views may represent the foundation with which people use to determine their preferences for redistributive policies. Preferences towards redistributive policies is also a more specific phenomenon in that these may take on different meaning depending on societal contexts. For instance, preferences for redistributive policies may mean something completely different for those in Sweden than they would for Americans.  One conclusion drawn from recent studies is that individuals have a basic understanding of income inequality (McCall and Manza 2011; McCall 2013). Indeed, a strong majority of the public consistently agrees that differences in income in their country are too large. Public opinions not only provide a view of citizens’ attitudes towards inequality but may also represent an important determining factor in shaping social policies. If public opinion towards inequality remains relatively stable in times of 15  rising inequality, then the increasing discrepancy between what the public desires and what reality actually is may become evident and individuals may begin to demand redistributive measures to be put into place.  To summarize, studying perceptions of inequality is important for two reasons. First, if the public is unaware of growing inequality, it is unlikely that they will demand a state response to the issue. It is difficult to justify any kind of government intervention if most individuals do not recognize it. Second, the values which influence public perceptions of inequality have implications for the types of governmental policies people will contemplate in order to make economic outcomes more equitable (Franko 2017; Newman, Shah and Lauterbach 2018). For instance, when pondering income inequality, people may take into account a number of relevant factors: expanding top income earners, growth of the middle class or general measures of the income distribution.  Studies exploring the interplay between relational classes and perceptions of inequality continue to be sparse. When the mechanisms rooted in class relations, understood in a Weberian sense, have been empirically tested, it has predominantly been on voting behaviour (Evans and Tilley 2012b), subjective self-placement (Haddon 2015), or social mobility (Erikson and Goldthorpe 1992) and not on perceptions of inequality. Those that do look at perceptions of inequality tend to use aspects of status as proxies for class and claim that if a significant relationship is found, then class matters in terms of people’s perceptions of inequality (McCall and Manza 2011; McCall 2013). As we know from Weber ([1922] 1978), classes are organized in terms of relations of production and the acquisition of wealth; whereas status groups are constructed and organized by principles of consumption and particular styles of life. Conflating class with aspects of status sunders the notion of distinct and cohesive relational class groups and leaves the importance of class open to scrutiny.  16  1.5. Outline of the Dissertation 1.5.1. Connecting the Empirical Chapters  Chapter two and Chapter three explore the relationship between social class and inequality views. Each chapter also reveals the important role that political persuasions play in this relationship. Chapter two sets the stage for my dissertation by revealing the importance of social class for viewing inequality critically. Chapter three builds on Chapter two by documenting how the relationship between class, political ideology and perceptions of inequality plays out during a recession. Both chapters explore whether social class has declined in shaping critical views towards inequality as death of class theorists assert. Contrary to these claims, the chapters reveal a significant relationship between social class and inequality views and outlines the trends in this occurrence.  Chapter two and three also assess the importance of politics, revealing the interplay between political ideology and individual economic conditions in shaping critical inequality views. Although these chapters are important in showcasing the role that class and politics play in viewing inequality critically, they do not explore the implications of societal-level factors in these orientations. To fill this gap, in Chapter four, I examine whether aspects at the country-level influence views towards inequality and how these coalesce with class. In part motivated by the lack of empirical research exploring the interplay between individual characteristics and broader factors in determining preferences towards inequality, this chapter illustrates that, depending on context, discontent with current levels of inequality differs depending on individual economic positioning. 1.5.2. Summary of the Findings  I now shift to a summary of my empirical results. Chapter two investigates whether class and socioeconomic status structure displeasure towards inequality differently, and whether these differences are patterned by political persuasion. This chapter explains how critical views towards inequality reflects class advantages; compares the degree to which class and status impacts views towards inequality; and examines how the relationship between class and attitudes towards inequality are complicated by political 17  preferences. In doing so, I reveal the need to separate social class and status when studying views towards inequality and illustrate the impact that political ideology plays in orienting inequality views.  This chapter also provides new theoretical insight. Contrary to common conjecture, this new empirical evidence calls into question the claim that factors based on the social relations of production play a reduced role in shaping attitudes towards inequality (Beck and Beck-Gernsheim 2002; Inglehart and Flanagan 1987; Inglehart 1990; Pakulski and Waters [1996] 2008). Chapter two also reveals that there is substantial class variation in the magnitude of tolerance towards inequality among those who are conservative. This is because the self-interests associated with class largely determines the views of those who are right-leaning, with working class conservatives viewing inequality more critically compared to their counterparts in more advantaged positions.  Chapter three is concerned with the role that class and politics play in shaping inequality views during recessions. This chapter asks two questions: 1) What motivates individuals to think critically about inequality following a major economic shock? 2) Are critical perceptions of inequality disparate across class and party and how are these shaped by The Great Recession? Here, I explore whether the various classes respond similarly or differently to an economic shock and how this is contoured by political ideology. While some studies have explored public opinions towards inequality amidst recessions (Cavaillé and Trump 2015; O'Connor 2017; Trump 2017), no research has simultaneously considered how class-polarization in critical views towards inequality is affected by party affiliation and whether this changes during an economic recession. To my knowledge, this is the first paper to do so. Chapter three concludes that there are patterns of homogeneity in how individuals view inequality, indicating that responses to TGR are patterned by social position. This chapter also reveals that following The Great Recession, the working class show little political variation in their views towards inequality; whereas the “upper class” showed more. Chapter four builds on the first two empirical chapters by exploring the relationship between social class positioning and national economic conditions on attitudes towards inequality. In this analysis, 18  I explore whether people are more critical of inequality in contexts when the actual reality of inequality is more visible. I also follow the common theme laced throughout my dissertation that social class has a strong role to play in moulding inequality views.  The main finding in Chapter four concerns the “upper class”, and whether members of this class view inequality as illegitimate in contexts when its effects are most visible. At low levels of inequality, the class differences in viewing inequality criticality are relatively large. Members of the “upper class”, are less critical of inequality under these conditions, perhaps because they feel inequality is justified as they are less aware of the inequality that surrounds them. On the other end of the class spectrum, the working classes are well aware of the effects of inequality and remain more resentful towards it. The working classes are mindful of the negative aspects connected to inequality in both unequal and more equal societies. As inequality increases, however, the views that the various classes have towards inequality begin to converge. In fact, in contexts where inequality is high, “upper class” inequality views are more critical than the working class.     19  Chapter 2: How Class Shapes Critical Views Towards Inequality: The Competing Forces of Stratification and Politics 2.1. Introduction Class analysis is under attack (Wodtke 2017). Scholars holding a “death of class perspective”, contend that the kind of class analysis founded on the social relations of production as in Marxist and Weberian scholarship is no longer important for understanding life chances and world views (Kingston 2000). Instead, lifestyles premised on a variety of identities are seen as increasingly key determinants of life chances and political attitudes (Beck and Beck-Gernsheim 2002; Inglehart 1990; Kingston 2000; Pakulski and Waters 1996).  According to this perspective, the class basis of society has fragmented. In its place has emerged new status group affiliations founded on a variety of identities, cultural practices, and values which now primarily determine life chances and political attitudes (Giddens 1991; Nisbet 1959). For example, marriage patterns tend to follow status divisions in terms of educational and cultural dimensions rather than class divisions; whereas post-materialist values have ushered in concerns with quality of life, civil liberties, and self-actualization that move beyond the traditional class-based cleavages of the past.  Value preferences around economic issues, such as those encapsulated in attitudes towards the legitimacy of inequality, are precisely what we might expect to be shaped most by social class (Andersen and Yaish 2012). Yet, even here, scholars arguing that class no longer matters posit that differences in material interests associated with class have diminished to the point where it is no longer a determining factor in shaping value preferences (Beck and Beck-Gernsheim 2002; Inglehart 1990; Pakulski and Waters [1996] 2008). Inglehart’s (1971; 1990) post-materialist thesis, for instance, originally purported that the rising prosperity that has accompanied modernisation and economic development has meant that factors based on the social relations of production play a reduced role in shaping attitudes, ushering in post-materialist values. Because basic economic needs are being met, economic issues are thought to be less intense and are assigned a relatively lower priority than in the past (Inglehart and Flanagan 1987). 20  Economic factors as a driving force of individual attitudes, which were useful during the early stages of industrialization, has seen their value diminish as scarcity has reduced.  Disparate findings with respect to research on the relationship between class and attitudes towards inequality may reflect differences in how class is operationalized in empirical studies, with many scholars conflating important theoretical distinctions between class and socioeconomic status. Scholars find that socioeconomic status and social class all have slightly different effects on and associations with various aspects of social life, such as economic security, prospects, and politics (Chan and Goldthorpe 2005; Chan and Goldthorpe 2007a; Chan and Goldthorpe 2007b; Chan and Goldthorpe 2007c; Chan, Birkelund, Aas et al. 2011; Goldthorpe 2010; Ridgeway 2014; Torssander and Erikson 2010). However, this distinction between class and socioeconomic status has often been ignored by sociologists when studying perceptions of inequality.  Although several studies have explored the individual attributes that underpin attitudinal differences towards inequality, there is a paucity of literature that has explored the differences situated firmly in class-relations (Im 2014). Rarer still are studies of critical appraisals of inequality using measures of both social class and socioeconomic status. Those that do exist (see Hadler 2005; McCall and Manza 2011) tend to focus on single countries and/or time periods, making it difficult to ascertain the generality of patterns cross-nationally (for an exception see Svallfors 2006). This study looks at broader relationships between class and perceptions of inequality across countries, which adds insights not available when focusing in on one or two countries.  Critical views towards inequality may be shaped not only by the different forms of stratification but also by politics. As an alternative to class and status, some analysts have shown the explanatory power of party affiliation in explaining perceptions of inequality (De Vries, Hobolt and Tilley 2018; Evans and Andersen 2006; Evans and Pickup 2010; Pickup and Evans 2013). This makes sense because dissatisfaction with income inequality is typically what we think of as a political left-right issue. For example, conservatives tend to be less critical of inequality because this is viewed as being associated 21  with individual rights, abilities and responsibilities. In this way, disapproval towards inequality relates to placement in terms of class and status, but also with politics (Singer 2011; van der Brug, van der Eijk and Franklin 2007; Weatherford 1982).  Contemporary accounts of peoples’ perceptions of inequality, however, often draw on insights from theories related to either aspects of stratification or from political affiliation, but do not consider the interplay between these. Working class individuals who vote conservative tend to be voting against their economic self-interest in that right-wing political parties tend to favour tax cuts for the rich, cutting benefits for the poor, and opposing redistribution. Along with Norris, Inglehart (2017; 2019) has recently proposed a “cultural backlash” to account for the rise of right-wing populism among the working class. They suggest that a backlash against cultural change by previous generations has launched a period where authoritarian populism is increasing. Nevertheless, they maintain that the rise of post-materialist values has “neutralized” class-based politics.  Left-leaners, on the other hand, tend to favour redistribution and reducing inequality regardless of their class position and so should be more critical of inequality based on their political persuasion. As Brooks and Manza (1997a; 1997b)  have shown, a person who belongs to a professional class is less likely to disapprove of inequalities based on this class position; yet, if they tend to be left leaning then disapproval may rise and they may reject existing levels of inequality even though their class position suggests otherwise.  Have interpretations of society founded on the social relations of production declined throughout the industrialized world as death of class theorists contend? Does class analysis continue to have purchase despite criticism? How might educational attainment and political orientation impact the relationship between class and attitudes towards inequality? Following Weber, this study introduces a clear conceptual distinction between social class, grounded in social relations of labour markets and production units, and Stand (status), situated in the structure of perceived social superiority and inferiority among individuals; primarily expressed through association and distinctive lifestyles. Using this framework, I draw on 22  International Social Survey Programme (ISSP) data from 26 countries to investigate how relational class distinctions (defined in terms of employment relations) correlate with perceptions of inequality. In doing so, I aim first to uncover the degree to which critical appraisals of income inequality are tied to social class in contemporary societies. Second, I compare the relationship between class, status and people’s perceptions of inequality to uncover whether each operates through different mechanisms. Third, I investigate whether the relationship between social class and attitudes towards inequality varies by political preference. 2.2. Theorizing Class, Education, and Attitudes Towards Inequality  A major issue for those claiming the demise of class relates to how class has been conceptualized. Class is one of the most frequently used concepts in sociology; however, it is also the most inconsistently defined. Indeed, even researchers who focus primarily on the study of class are often referring to quite different concepts (McCall 2013). Empirical research on public opinions towards inequality is no exception, with social class defined inconsistently and often in ways that conflate it with aspects of socioeconomic status. For instance, McCall and Manza (2011) use income as a proxy for class, finding that individuals in lower income groups are significantly more supportive of government redistribution than those in higher income groups. Dimick et al. (2017; 2018) focus on wealth as a measure of class, finding that preferences towards redistribution are dependent on individual wealth, but with the caveat that this varies by the actual level of inequality in a society. In other work, McCall (2013) uses education to operationalize class, finding that people with more education are likely to favour less inequality. Conversely, Koçer and van de Werfhorst (2012) find that higher educational attainment is associated with favourable perceptions of income inequalities. The divergent findings on the operationalization of class point towards a conceptual artefact that may be at play. While often used as a proxy for class, income leaves out important elements that may be critical for understanding the relationship between class and attitudes towards inequality. Likewise, while empirically related, class and education remain conceptually distinct, and may impact ideas about 23  inequality in different ways. Income level may obviously shape peoples’ concerns with income inequality. Those with higher incomes have less reason to be concerned about inequality, and its effects, than those with less. However, although the income distribution is an important aspect of stratification which is central to theories of class structure, it does not capture fully the mechanisms of the class structure (see Zhou and Wodtke 2018). The broader social relations of production also matter. For instance, employers of those classes encompassed under a “service relationship” typically have stronger incentives to retain such workers, increasing their job security, and reducing the variability of their income flows. Compensation schemes also may have elements of deferred compensation, with stronger wage growth over time as a result (ibid). Large employers not only typically have high incomes, but also have a great deal of control over decisions in terms of their own compensation and income distributions in their firms, which may result in particularly strong feelings that overall income distributions in their country are fair.  Those in more advantaged employment relations are also more likely to enjoy a “social wage”, which includes various social protections with their compensation schemes whereby employers provide broader benefits (e.g. improved health/dental insurance, paid holidays and leave), such that direct employment income is not the only important aspect of their compensation (Cranford and Vosko 2006). Even if they have relatively low incomes early in their careers, they may be less critical of inequality as they expect to be relatively advantaged later in the future and as they are less exposed to risks that they would otherwise need a higher income to buffer.  Conversely, the more disadvantaged classes, those encompassed by the “labour contract” have positions which are characterized by a specific exchange of money for effort and limited protection from risk as well as lower wages (Bengtsson, Berglund and Oskarson 2013). Individuals in different class positions thus differ not only in terms of income level, and these other aspects of their class position may contribute to their opinions towards inequality. The key advantage of a class-based measure is that the various class categories capture these broader elements in addition to the income distribution. 24  Class position and education also overlap, albeit imperfectly. They may have a similar empirical relationship towards inequality views insofar as they both reflect resources that tend to elevate one’s social standing, and hence potentially dampen concern with inequality. However, they are not identical and may have distinct effects. Classes are organized in terms of relations of production; whereas status groups are constructed and organized by principles of consumption and particular styles of life. Both may shape perceptions of inequality, but differently (Hadler 2005; McCall 2013). For instance, knowledge about the level and extent of inequality in society may be more accessible to those with higher education and thus highly educated people may hold more critical views towards inequality even though their generally advantaged class position otherwise gives them a stronger stake in the status quo. Or perhaps, those who are highly educated may feel even more strongly that their class position is an achieved status reflective of merit, and hence be even less likely to negatively judge the income differences in their societies (Jackman and Muha 1984). Education may itself impact attitudes towards inequality in a way conceptually distinct from the relationship between class and inequality views. How much of the influence of social position on perceptions of inequality is explained by differences in education between individuals of different classes? Opinions towards inequality might be affected by education, yet those in more advantaged classes typically have higher levels of education. There is clearly a strong connection between social class and education; people with higher levels of education typically end up in more advantaged classes (Goldthorpe 1996). Education is often a precursor to class. As a result, a portion of the association between class position and inequality views may reflect the prior effect of education insofar as people with higher levels of education end up in more advantaged classes. This illustrates the need to separate aspects of status and class within an analysis as both are tapping into differing yet somewhat correlated systems of stratification (Chan and Goldthorpe 2007a; Chan et al. 2011; Goldthorpe 2010; Torssander and Erikson 2010).   Whether the content of class is captured through mechanisms of exploitation (Wright 1997) or by employment relations (Erikson and Goldthorpe 1992), class analysis seeks to offer not only an 25  explanation but also a theoretical description of the sources of inequality. As the literature points towards competing findings regarding the relationship between class-related aspects and inequality views, it seems reasonable to complement the concept of class with other distinct aspects of stratification (Andersen and Curtis 2012; Curtis 2016; Hadler 2005; McCall and Manza 2011; McCall 2013). Social class and status may not be reducible to each other, which means that in order to advance the study of peoples’ views towards inequality, the stratification of different views should be investigated by harnessing reliable measures of both, rather than a single measure that conflates the two. This chapter seeks to disentangle the class and status components of stratification that affect individual opinions towards inequality.  2.3. Class and Opposition to Inequality Research shows that there are social class differences in the distribution of income (Piketty 2014). In the United States, for example, there have been increasing income differences between social classes, primarily driven by growing income for high-level managers and large proprietors alongside a reduction in wages for workers and independent producers (Wodtke 2016). It seems reasonable to expect that these growing material differences between classes may lead to similar gaps in the classed perceptions of these differences.  Individuals who occupy different positions within the social relations of production might be expected to behave and think in differing ways when it comes to income inequality. Assuming this to be the case, I propose that individuals residing in different class locations develop perceptions of society that are relatively consistent with the interests of the class that is objectively theirs. While not all members of a particular class location will share the same perceptions, the probability that they will share perceptions which are consistent with the objective interests connected to their class locations will be higher than for members of other class locations. While individual attributes may affect the strength of the association between the class structure and subjective inequality, this association will be based on the class structure rather than on personal attributes of the individual. 26  The various forms of compensation regulated by the different contractual agreements of the various classes also act as protections from reductions in income and provides a strong rationale for a relationship between class position and inequality views. It is reasonable to suspect that the disadvantaged compensation characteristic of those classes regulated by the labour contract (including limited career prospects, specific exchange of money for effort on a piece or time basis and minimal job security) may lead them to be more critical of income inequality; while the more advantaged compensation package of those classes regulated by the service relationship (better employment and payment conditions, potential promotional prospects and occupational security), may lead them to be less critical of income differences.  Individuals are likely to favour economic equality and redistribution if they are to benefit from it (Meltzer and Richard 1981). Previous research has suggested that those from more advantaged classes tend to be more accepting of inequality because they are less affected by the negative consequences of it, and in turn, tend to lose the most from redistributive policies (Meltzer and Richard 1981). As higher managers and professionals are the most advantaged class with regards to occupational security, employment and payment conditions, and potential promotional prospects, they have less reason to be critical of inequalities and preserve the status quo. Large employers are likewise materially advantaged, and also have considerable control over the structure of inequality in their own firms, also suggesting greater investment in the status quo. It seems reasonable then to compare the inequality views of the other classes to the class which I suspect to be the least critical of inequality.  On the other end of the class structure, those from the more disadvantaged classes tend to be more critical of inequalities because they stand to gain more if redistributive policies are put into place (Andersen and Heath 2002). The disadvantages of the labour contract—short-term exchange of money for effort—can explain why members of the routine classes may be more critical of income inequalities and support redistributive programs that distribute wealth more equally (Goldthorpe, Lockwood, Bechhofer et al. 1970). The working classes (routine, lower technical,) should be more critical of income inequality in comparison to Large employers/higher managers and professionals because their members have 27  comparatively lower wages and limited occupational security and may therefore have strong incentives for collective action leading to a state response (Kalleberg 2000). Thus, it is in their interest to be more critical of current levels of inequality as this may lead to the formation of policies which are beneficial to them. The following hypothesis emerges from the above theory: Hypothesis 1: The “routine” and “lower technical” classes are more critical of income inequalities in comparison to “Large employers/higher managers”.  2.4. The Interplay between Class, Education and Attitudes Towards Inequality  Research finds that higher levels of education are associated with more critical inequality views (Hadler 2005; Koçer and van de Werfhorst 2012; McCall 2013). The rationale behind this is that higher levels of education may produce more liberal values that undermine a belief of a natural social order or one that is premised on meritocracy (Hadler 2005; Koçer and van de Werfhorst 2012; McCall 2013). Increased education may cultivate values more conducive to the emergence of opinions which are in favour of redistribution, which is often sought after when one is aware of inequalities and considers these unjust. Education is thought to broaden individual horizons which gives people the ability to perceive social issues differently and thus opens up the possibility that they may be more critical of political issues (Bengtsson et al. 2013).  Conversely, an alternative theory known as the ideological refinement perspective, regards education not as generating values conducive to progressive politics, instead advanced education provides heightened intellectual training and a more refined mastery of information which enables those who are highly educated to further their self-interests more effectively (Jackman and Muha 1984). Those advantaged in terms of education may hold status beliefs favouring ideologies which legitimize and justify inequality and so may be less critical of inequality. To test these competing theories, I pose the following hypothesis: Hypothesis 2: As education rises people are increasingly likely to view inequality critically. 28  Class and education are distinct components of stratification and their relationships with viewing inequality may be different. Moreover, it is possible that education moderates the relationship between class and views towards inequality. In particular, class may be more salient where other dimensions of status are relatively lacking, such as among those with lower levels of education. Education, conversely, may matter more in relatively advantaged classes such as the lower professional/technical class where members may be highly educated in fields of study where attention is focused on teaching or caring for individuals (e.g. nurses, teachers)(Werfhorst and Graaf 2004). This would tend to increase any liberalizing effect of education. This leads to the next hypothesis: Hypothesis 3: Education moderates the relationship between class and critical views towards inequality: That is, as education rises the impact of class on perceptions of inequality will be weaker.  2.5. Class and Party In addition to theories based on class and status, extensive literature has demonstrated that core values associated with individual vote choice shape orientations towards economic and political equality (Andersen and Heath 2002; Caínzos and Voces 2010; Evans 2002; Hicks, Jacobs and Matthews 2016; Sosnaud, Brady and Frenk 2013; van der Waal, Achterberg and Houtman 2007). Party affiliation appears to play a role in shaping individual judgements towards inequality. Inequality tends to increase under conservative parties and the rhetoric espoused by these parties may shape public opinion towards more favourable views towards it (Evans and Tilley 2017). Those favouring conservatism are likely to view the current redistribution of material resources as legitimate in their society (Kelly and Enns 2010). Conversely, those holding political preferences characteristic of the left, which view economic inequality as unjust, tend to agree that income inequalities are too large (Hadler 2005). There is debate as to the direction of the relationship between party affiliation and economic perceptions (Evans and Pickup 2010; Pickup and Evans 2013). Recent research, however, shows that party affiliation may be a key driver of economic perceptions rather than the reverse. (De Vries et al. 29  2018; Evans and Andersen 2006; Evans and Pickup 2010; Pickup and Evans 2013). For example, Evans and Pickup (2013) conducted an analysis of the political conditioning of economic perceptions in the 2004 U.S. presidential election using panel data from 2000, 2002, and 2004. Using two-stage least-squares estimation, they find that economic perceptions have “no discernable impact on vote, presidential approval and party identification” (Evans and Pickup 2010:1238). Moreover, Evans and Andersen (2006) find that retrospective economic perceptions are strongly influenced by prior views towards the incumbent party. They examined the British Election Panel Study across the 1992-1997 British electoral cycle and found that the effects of economic perceptions are often substantially overestimated and that the effects of party support on economic perceptions are much stronger. By taking into account the long-term conditioning effects of party support on economic perceptions, they contend that “the direction of influence between economic perceptions and political preferences is disproportionately from politics to economics rather than vice versa” (Evans and Andersen 2006:196). There is, then, firm scholarly support for considering party affiliation as a predictor of economic perceptions.1 This implies that there may be a relationship between party affiliation and views towards inequality.    Not only may there be an independent relationship between party affiliation and inequality views, but this may be moderated by class. For instance, right party affiliates in routine classes may be more critical of income inequalities because the characteristics of their class position (e.g. limited security, increased precarity, low wages) weakens their conservative perspective, which indicates that they should be more tolerant of inequality. Conversely, party affiliation may matter differently across advantaged class categories. A member of the large employers, higher managers and professional class who affiliates with the left may reject inequality, even though their generally advantaged class position suggests the opposite. For this segment, their left-leaning political ideology towards society may weaken the                                                      1I recognize that there is much debate between whether political ideology impacts inequality views or the reverse and cannot be certain that my consideration of the direction of the relationship is correct. Further, it is certainly possible that political ideology could be affected by class and thus act as a mediator between the relationship between class and inequality attitudes. As I do not consider this aspect, this should be considered a limitation of this research. 30  relationship between their advantaged class position and acceptance income inequalities. What these scenarios illustrate is that right or left party affiliation may reduce the impact of class on inequality views. As a result, I pose the following hypothesis:  Hypothesis 4: Party affiliation moderates class effects on perceptions of inequality  2.6. Methods 2.6.1. Data To test these hypotheses, I use data from the International Social Survey Programme (ISSP). The dataset relevant for this article is the one on social inequality which has been used in four waves: 1987, 1992, 1999, and 2009. The survey wave 1987 only has data available for 8 countries (Austria, Australia, Hungary, Italy, Poland, Switzerland, United Kingdom and the United States). To maintain generality of patterns cross-nationally I elected to remove the survey wave 1987. The most recent data in this dissertation are from 2009, which means that the results do not pertain to the present day. Nevertheless, the ISSP offers the best available source of cross-national data on public opinion as it relates to perceptions of inequality. There are other sources of data that look at public opinion towards income inequality (see Kenworthy and McCall 2008); however, I use data from the ISSP because it covers a greater number of countries, is more comparable across countries, and is more effective at tapping into people’s opinions towards inequality. The ISSP thus allows a more general test of the relationships at interest than data restricted to a smaller number of countries or time periods.  Further, to maintain comparability across models, I drop observations with missing values on any variable used in any of the models. Approximately, 21 per cent of data are missing from the class variable due to non-responses on the occupation component. A further 2 per cent are missing either because respondents never had a job, are not currently working, are members of the armed force or retired. Education has roughly 1 per cent of cases missing. The largest number of missing values pertains to the party affiliation variable. In total, 49 per cent are missing because respondents either did not vote in their 31  country’s last national election (34 per cent) or the respondent’s political preferences could not be specified as either left, right or centre (15 per cent). Excluding those with missing values on any covariate reduces the observations from 88, 167 to 33, 678. While such listwise deletion can bias coefficients when data is not missing at random, there is insufficient information to reasonably use imputation strategies (which often rely upon untestable assumptions).  The impact of missing party support values means that results only pertain to those who are politically engaged. Appendix A provides details on the relationship between missing values on party affiliation and key covariates to provide context on possible implications of this omission. Overall, the correlation between the variables used in the analysis and the party support variable were low (r = .05 to r=.19), which may indicate that this is unlikely to bias estimates of key covariates (Mustillo 2012). Nonetheless, as a sensitivity analysis, I performed a Wald test to formally test whether including those with missing values on the party support variable (coding them as additional categories for party support) affects estimates of other covariates. In terms of the key variables, omitting those who are not politically engaged biases the estimates for the main effects of the intermediate and routine classes as well as those with less than secondary education. This suggests that the estimates for these groups are restricted to those who are politically active. Results for this strategy can be found in Table A1 in Appendix A.  2.6.2. Predictors The composite measurement of class employed in this analysis is inspired by the Weberian tradition. Developed by Rose and Harrison (2007; 2010) in conjunction with Eurostat, the European Socio-economic Classification (ESeC) represents a measure of class based conceptually on the EGP (Erikson-Goldthorpe-Portocarero) schema. The aim of the class schema is to differentiate social positions on the basis of labour markets and production units by way of the employment relations that these entail. The employment relationship is determined firstly by the distinction between employers, employees, and self-employed; which is similar to the classes understood by both Marxists and Weberian interpretations 32  of class (Erikson and Goldthorpe 1992; Goldthorpe et al. 1970; Wright and Perrone 1977; Wright 1976; Wright 1997).  The first step when creating the employment status component is to ascertain whether the respondent was employed or self-employed. The second step involves obtaining the supervisory status of the respondent. This differentiation is made because higher supervisory occupations have weaker forms of the service relationship in terms of skills, expertise, and knowledge, yet are regarded as having superior employment contracts compared to those who are supervised. Once employment status is known, the final step is to assign individuals to their respective classes depending on their occupation. Individual occupation is collected by asking respondents to list their occupation which, if it not done already, is then coded using the ISCO-88 3 code classification. Within the category of employees, a further distinction is made between those involved in a service relationship with their employer and those regulated by a labour contract. Service relationships are characterised by a more varied exchange which includes not only compensation for work done through salary but also potentially important aspects which include salary increments following scales, employment security and pensions rights, and the possibility of career advancement. This is contrasted with the labour contract which involves a relatively short-term exchange of income for labour. Workers under the labour contract exert effort under the supervisions of employers or hired agents in return for wages that are calculated by piece or on a time basis. Class compositions by country are shown in Figure 2.1.33  Figure 2.1 Class compositions for each country (proportions)  34  Education is measured categorically by qualification obtained, using a slightly re-arranged subset of the International Standard Classification of Education (ISCED-97). It would have been beneficial to include income within the models as this would have allowed me to explore how it may affect the relationship between class and perceptions towards inequality. Indeed, qualitative research has documented that differences in material well-being and income are important aspects considered by individuals when thinking about inequality (Irwin 2018). However, income data was not collected in comparable ways across countries. Information varies widely, for example, whether the survey asked the respondent for income per month or year, before or after tax and in what currency. Nevertheless, the extent to which income moderates the relationship between class and inequality views is certainly something that should be explored in future research in this area. I relied on the decision of those who disseminated the data who coded individuals from a particular country as either “left”, “right” or “centre”. This is important insofar as local knowledge is crucial for properly classifying particular political parties. For instance, in the United States those who consider themselves “Strong Democrats” or “Not very strong Democrat” are considered centre, left; “Independent, close Democrat”, “Independent”, or “Independent, close Republican” regarded as centre, liberal; and “Not very strong Republican, “Strong Republican” coded as right, conservative. While not perfect, it is my contention that the party affiliation variable used in this analysis does have some usefulness in determining, generally speaking, where individual political preferences can lie on the partisan spectrum as either “left”, “centre” or “right”. I recoded the party support variable such that “Far left”, “left, centre left” are considered “left” (1); “Centre, liberal” as “centre” (2); and “Right, conservative, “Far right” as “right” (3). For the analysis, “centre” is the reference group. Those with no party preference, or “other no specification” are coded as missing. 2.6.3. Outcome  The dependent variable has been recoded from the following survey question: “How much do you agree or disagree with the statement ‘Differences in income in [respondent’s country] are too large?’”. 35  Response categories include: strongly agree (1), agree (2), neither agree nor disagree (3), disagree (4), and strongly disagree (5). Following previous research, I recoded the variable into an inequality index with the following values: ‘‘agreed strongly’’5, ‘‘agree’’ 4, ‘‘neither agree nor disagree’’ 3, ‘‘disagree’’ 2 and ‘‘disagree strongly’’ 1.  Strictly speaking, this continuous approximation of an ordinal measure violates the fundamental assumptions of an ordinal level measure (Johnson and Creech 1983). Computing the means and standard deviations of Likert scale measurement is often considered inappropriate; instead, nonparametric statistics should be used. There are others who contend that treating Likert scales as continuous is beneficial as they tend to operate as roughly linear in practice (Harwell and Gatti 2001; Sullivan and Artino 2013; Wu and Leung 2017). While I cannot make inferences about the underlying characteristics of the Likert scale used to measure views towards inequality, this should not imply that the conclusions drawn from the statistics from this scale are invalid (Norman 2010). Parametric tests not only can be used with ordinal dependent variables, but parametric tests on such variables are often more robust than nonparametric tests. As Sullivan (2013: 542) notes, parametric tests on Likert scale responses are often found to be “sufficiently robust to yield largely unbiased answers that are acceptably close to “the truth””. Indeed, it is common in empirical studies examining views towards inequality, to treat Likert scales with five or more categories as continuous (Evans and Kelley 2018; Larsen 2016; Roex, Huijts and Sieben). As a sensitivity analysis, I modelled the ordinal version of the attitudes towards income inequality variable using ordered logistic regression. The results of this modelling strategy are displayed in Table A2 in the Appendix. The results revealed similar patterns in the relationship between class, education, party support and their interactions compared to when the dependent variable is treated as a continuous measure.  Note that, this survey item does not directly measure how much inequality there is. What it does tap into is the awareness of the degree of inequality and the associated degree of fairness about the perceived level of inequality. It is a complex question because it combines two aspects: the perceptions of inequality and the opinions towards fairness of inequality. People need to see inequality in their society 36  first and also hold critical inequality views in order to justify a possible state response. In this way, inequality views represent the foundation with which people use to determine their preferences for redistributive policies.  2.6.4. Additional Controls  Although my interest lies predominantly in exploring class and status effects on subjective inequality, all models also control for gender, age, marital status and religion (Hadler 2005; McCall and Manza 2011; Svallfors 2005). Scholars have identified a gender gap in preferences for government spending, which is likely related to attitudes towards inequality, justifying its inclusion as a control in the models (Dodson 2007; Edlund 2003). The marital status variable was dichotomized such that married is equal to 1 while widowed, divorced, separated, never married/single are considered zero. Common law individuals are considered single/never married. I include religion in the analysis as a control as it has been found that there is a correlation between attitudes towards inequality and religion (Jordan 2016). Roman Catholic is selected as the reference category.  I do not factor into the models possible antecedent conditions which could affect the relationship between class and inequality views. For instance, parental class is an important antecedent condition that could affect both current class and views towards inequality. The ISSP dataset does contain a variable on parental occupation which I could have arranged into a measure of class using a “simpler” version of the ESeC class schema. There were, however, a high number of missing values for various countries used in the analysis. For example, values were only available for Bulgaria in 2009. Philippines was missing values for 1992 and 1999. Slovenia was missing values for 1992 and the U.K. was missing values for all years. Descriptive statistics for the key variables by country are shown in Table 2.1   37  Table 2.1 Descriptive statistics for each country  Mean subjective inequality Proportion with at least a university degree Proportion Voting Right Proportion Voting Centre Australia 3.75 .22 .47 .05 Austria 4.21 .08 .34 .15 Bulgaria .4.70 .14 .38 .19 Canada 3.79 .30 .12 .78 Chile 4.25 .08 .25 .31 Cyprus 3.69 .25 .45 .23 Czech Republic 4.27 .10 .39 .22 France 4.50 .26 .22 .33 Germany 4.22 .05 .32 .08 Hungary 4.49 .05 .43 .06 Italy 4.49 .14 32 .15 Japan  .4.01 .18 .51 .39 Latvia 4.52 .18 .16 .38 New Zealand 3.83 .17 .64 - Norway  3.69 .23 .37 .20 Philippines 3.48 .16 .30 .58 Poland 4.27 .12 .28 .12 Portugal 4.65 .09 .05 .34 Russia 4.54 .21 .23 .33 Slovak Republic 4.54 .11 .17 .47 Slovenia 4.38 .10 .26 .52 Spain 4.20 .12 .21 .29 Sweden 3.83 .22 .32 .17 Switzerland 4.20 .14 .51 .07 UK 4.03 .14 .41 .16 US 3.81 .25 .26 .45  38  2.7. Analytical Strategy  The multilevel dataset I am working with has three levels: level one refers to the individual observations; level two refers to country-year; and level three country. In order to adjudicate between whether to employ fixed effects (FE) or random effects (RE) for the first two empirical chapters, I ran a Hausman test. The Hausman test has been devised to test whether the results from a FE and RE model are systemically different from one another. If there are significant differences between coefficients using FE and RE then it is better to use FE. I first specify the dataset as panel at the country-year level as the same countries were surveyed at different time periods. I specify the null hypothesis that the preferred model is the RE versus the alternative, FE. A large and significant Hausman statistics would indicate large and significant difference between the two sets of coefficients of the FE and RE models. In this case I reject the null hypothesis that the two methods are acceptable, instead favouring the alternative hypothesis that the FE is preferable. The test revealed that there are significant differences between the two sets of coefficients (p < .001), so I can reject the RE model and apply a FE model.  The Hausman test provides statistical support for a FE model. The substantive focus in this chapter also supports this strategy. There are advantages of the FE model when the research interest is focused on looking at individual-level processes. As my interest in this chapter focuses in on exploring the relationship between individual-level processes and inequality views, the FE is the beneficial modelling strategy as it “controls out” any higher levels processes, which are distinctive to higher-level units (Bell and Jones 2014:139; Schmidt-Catran and Fairbrother 2015). This approach enables me to explore the relationship between individual-level variables net of any higher-level attributes within particular countries.  As I treat my dependent variable as an interval-level measurement I employ linear fixed effects regression models. In the fixed effects models, all country-level variables are controlled out by using country-fixed effects as dummy variables. I also include year fixed effects in order to capture the influence of time trends. The first model estimates class differences in perceptions of inequality alongside 39  possible confounders. This model tests hypothesis 1 which suspects that subjective inequality varies according to the ESeC class schema. In model 2, I test for the possible impact of education on critical inequality views, net of additional controls in order to address hypothesis 2. This model explores whether there is a relationship between higher levels of education and critical inequality views. Model 3 includes class and education together in order to determine whether the relationship between class and critical inequality views endures as education is included. Model 4 tests the claim made in hypothesis 3, that education moderates the relationship between class and critical inequality views. In model 5, I explore the relationship between party affiliation and inequality views before introducing the interaction with class. Hypothesis 4 presumes that there is a possible interaction between the class effects and party affiliation. Model 6 tests this claim. All models apply probability weights.  2.8. Results  Table 2.2 displays each of the six models. I begin by exploring the relationship between class and critical views towards inequality. I expect that the more disadvantaged classes, that is those classes characterized by the labour contract, are more critical of inequalities in relation to the more advantaged classes. The first column provides information for Model 1, providing a test relating to the influence of social class on subjective inequality. This model provides support for Hypothesis 1, demonstrating that the routine, lower technical, and lower sales and service classes—those class positions regulated by the labour contract—are more critical of income differences in their societies in comparison to large employers, higher managers/professionals. The significant positive coefficient values next to the various class categories indicate that all of these classes tend to be more critical of income inequalities than large employers, higher managers/professionals. For instance, the routine classes view inequalities .39 higher on the inequality scale than large employers, higher managers/professionals.  Looking at the “contrasts” column, which reports where each class category differs significantly from others, we see that both the lower technical and routine classes are also significantly more critical of income inequality than Lower managers/professionals, higher supervisors/technicians; Intermediate 40  occupations; Small employers/self-employed; Lower supervisors/technicians; and Lower sales and service. Overall, class groups tend to differ from all others in their views towards inequality, save, occasionally, those closest to them in the overall hierarchy. The observed patterns suggest that there are strong class differences in preferences for inequality. Model 2 introduces educations as a measure of status. This model fails to support Hypothesis 2, which posited that as education rises individuals are more critical of inequality. In comparison to those with a university degree, those with less than secondary school qualifications actually view inequality .32 higher on the scale, whereas those with secondary qualifications are .24 higher. Indeed, looking at the “contrasts” column, those with less than secondary schooling are significantly more critical of income inequalities than those with higher levels of education, all else equal.  Model 3 includes both the class and education variables together. When education is included alongside class, the effects of education remain substantively the same but are weaker—in comparison to those with university degrees, those with less than secondary schooling are now .16 higher on the scale, representing a significant reduction of 50 per cent from the previous model (P < .001). The class contrasts are also somewhat smaller, although the general pattern holds. These results reveal evidence that education partly confounds the relationship between class and critical inequality views. To further investigate the complex interplay between class and education as predictors of attitudes towards inequality, model 4 includes an interaction term to test Hypothesis 3 as to whether education moderates the impact of class on perceptions of inequality. There is one significant interaction between Lower managers/professionals, higher supervisory/technicians with secondary/higher secondary educational qualifications, and jointly the interaction terms are statistically significant (p < .01) which indicates that the relationship between class and perceptions of inequality varies depending on education2. Table 2.3 provides the average marginal effects for the classes across levels of education. The                                                      2 Joint tests of significance for all variables across the models are provide in table A.3 in the Appendix. 41  table reveals that the majority of significant differences between classes are among those individuals who possess no more than secondary/higher secondary qualifications. Between classes with above higher secondary and university/graduate levels of education there are fewer significant class differences. These results confirm hypothesis 3 that education moderates the relationship between class and critical inequality views. Class seems to matter only among the less educated.  Model 5 introduces party affiliation alongside class. Left leaners are .15 times higher on the scale compared to centrists; while conservatives are .18 lower (P < .001). Jointly, the terms are also significant (P < .001). This confirms the common finding that those favouring conservatism are less critical of inequality; whereas left-leaners tend to be comparatively more critical. When included alongside party affiliation, the effects of class remain substantively the same but are weaker—in comparison to Large employers/higher managers and professionals routine workers are .27 higher on the scale, representing a significant reduction of 31 per cent from model 1 (P < .001). This suggests that part of the relationship between class and inequality views appears to be due to party affiliation.  Model 6 includes an interaction term between political affiliation and class to test Hypothesis 4 which posits that the relationship between class and inequality views are conditional on political orientations. Jointly, the interaction terms are statistically significant (p < .001) which indicates that the relationship between class and perceptions of inequality varies depending on an individual’s political persuasion. Table 2.4 provides the average marginal effects from model 6. The letters next to party affiliation indicate which classes are significantly different across the political spectrum. We can see that there are more significant differences between classes in those who are conservative and there are comparatively less in those who are centrists and even fewer in those who are left-leaners. Moreover, focusing in on conservatives, we can see that those classes in the “middle” of the class spectrum are not as different in their inequality views compared to the “upper” and “lower” classes. Among conservatives, the differences among classes in viewing inequality emerges most clearly at the opposite ends of the class structure.  42   Interpretation of the interactions from the results of model 6 alone is difficult. Figure 2.2 displays the linear predictions of inequality views according to class and party affiliation to more clearly display the relationship between class, vote choice, and subjective inequality. The figure reveals that those who are left leaning are critical of income inequalities regardless of their class position. There is slight linearity in the relationship between lower class positioning and more critical attitudes to inequality among those in the centre of the political spectrum. What this graph shows is that the aggregate association of disadvantaged class positioning and more critical attitudes towards inequality is driven largely by more conservative individuals. In those who are right leaning, there is considerable class differentiation, as seen in the differences between each class category. Large employers/higher managers who vote right are decidedly less critical of inequalities in their society compared to right leaning routine workers. This reveals that while those who affiliate with the right tend to be less critical of inequalities, the difference attenuates for the less advantaged. The ramifications of class in shaping people’s perspectives towards inequality are most clearly visible for conservatives.  43  Table 2.2. Fixed effects of linear regression models and interactions between class, education and political affiliation in critically inequality views  (Model 1) (Model 2) (Model 3) (Model 4) (Model 5) (Model 6) Contrasts Social Class (ref: Large employers/higher managers and professionals)        a Lower managers/professionals, higher supervisors/technicians 0.183***  0.157*** 0.194*** .144*** 0.078 b1, d1, e1, f1, g1/ b3, d3, e3, f3, g3/b5, d5, e5, f5, g5  (0.02)  (0.02) (0.03) (02) (0.04)  b Intermediate occupations 0.264***  0.214***  0.244*** .199*** 0.153*** a1, d1, e1, f1, g1/ a3, e3, f3, g3/a5, f5, g5  (0.02)  (0.02) (0.05) (.02) (0.05)  c Small employers/self-employed 0.233***   0.163*** 0.191 .173*** 0.144* d1, e1, f1, g1/ d3, e3, f3, g3/d5, e5, f5, g6  (0.02)  (0.03) (0.12) (.03) (0.06)  d Lower supervisors/technicians 0.319***   0.255*** 0.194** .240*** 0.227*** a1, b1, c1, f1, g1/ a3, c3, f3, g3/a5, c5, f5  (0.02)  (0.02) (0.06) (.02) (0.05)  e Lower sales and service 0.334***   0.264*** 0.279*** .224*** 0.183*** a1, b1, c1, f1, g1 / a3, b3, c3, f3, g3/a5, c5, f5  (0.02)  (0.02) (0.07) (.02) (0.05)  f Lower technical 0.396***   0.320*** 0.059 .288*** 0.178*** a1, b1, c1, d1, e1/ a3, b3, c3, d3, e3/a5, b5, c5, d5, e5  (0.02)  (0.02) (0.15) (.02) (0.05)  g Routine 0.388***   0.308*** 0.246* .266*** 0.189*** a1, b1, c1, d1, e1, / a3, b3, c3, d3, e3/a5, c5, c5,   (0.02)  (0.02) (0.11) (.02) (0.04)  Controls        Female 0.116*** 0.125*** 0..126*** 0.126*** .116*** 0.112***   (0.01) (0.01) (0.01) (0.01) (.01) (0.01)  Age 0.004*** 0.003*** 0.004*** 0.004*** .003*** 0.004***   (0.00) (0.00) (0.00) (0.00) (.00) (0.00)  Married -0.008 -0.016 -0.007 -0.008 -.002 -0.002   (0.01) (0.01) (0.01) (0.01) (.01) (0.01)  Not religious (ref: catholic) 0.098*** 0.096*** 0.103*** 0.104*** .057*** 0.052**   (0.01) (0.02) (0.02) (0.02) (.02) (0.02)  Protestant -0.009 -.011 -0.010 -0.009 -.004 -0.002   (0.02) (0.02) (0.02) (0.02) (.02) (0.02)  Other religion 0.076*** 0.080** 0.080 0.080** .045 0.044   (0.03) (0.03) (0.03) (0.03) (.03) (0.03)  1999 (ref: 1992) 0.083*** 0.098*** 0.092*** 0..092*** .093*** 0.093***   (0.02) (0.02) (0.02) (0.02) (.02) (0.02)  2009 0.044** 0.068*** 0.060*** 0.060*** .057*** 0.057**   (0.02) (0.02) (0.02) (0.02) (.02) (0.02)  Education (ref: University/Post-graduate studies)        h Less than secondary  0.318*** 0.160*** 0.42 .166*** 0.176*** i2, j2/ i3, j3/j5   (0.02) (0.02) (0.10) (.20) (0.02)  i Secondary/Higher secondary  0.240*** 0.128*** 0.154*** .136*** 0.143*** h2, j2/ h3, j3/j5   (0.02) (0.02) (0.04) (.02) (0.02)  j Above higher secondary  0.143*** 0.080*** 0.120* .096*** 0.103*** h2, i2/ h3, i3/h5, i5   (0.02) (0.02) (0.05) (.02) (0.02)  Class*Education Interactions (ref: Large employers/higher managers and        44  professionals with university/post-graduate credentials) Lower managers/professionals, higher supervisory/technicians*Less than secondary    0.085        (0.11)    Lower managers/professionals, higher supervisory/technicians*Secondary/Higher Secondary    -0.096*        (0.05)    Lower managers/professionals, higher supervisory/technicians*Above higher secondary    -0.032        (0.06)    Intermediate occupations*Less than secondary    0.005        (0.12)    Intermediate occupations*Secondary/Higher Secondary    -0.041        (0.06)    Intermediate occupations*Above higher secondary    -0.083        (0.07)    Small employers and self-employed*Less than secondary    -0.063        (0.16)    Small employers and self-employed*Secondary/Higher Secondary    -0.006        (0.13)    Small employers and self-employed*Above higher secondary    -0.061        (0.16)    Lower supervisors and technicians*Less than secondary    0.110        (0.12)    Lower supervisors and technicians*Secondary/Higher Secondary    0.061        (0.07)    Lower supervisors and technicians*Above higher secondary    -0.019        (0.09)    Lower sales and service *Less than secondary    0.096        (0.12)    Lower sales and service *Secondary/Higher Secondary    -0.051        (0.08)    Lower sales and service *Above higher secondary    -0.075        (0.10)    Lower technical*Less than secondary    0.269        (0.18)    Lower technical*Secondary/Higher Secondary    0.266        (0.16)    Lower technical*Above higher secondary    0.215        (0.17)    Routine*Less than secondary    0.071        (0.15)    Routine*Secondary/Higher Secondary    0.079        (0.12)    Routine*Above higher secondary    0.017        (0.13)    Party Affiliation (ref: Centre)        k Left     .154*** 0.232*** l5      (.01) (0.04)  l Right     -.182*** -0.395*** k5      (.02) (0.04)  Class*Party Interactions (ref: Large employers/higher managers and professionals        45  affiliating politically with the centre) Lower managers/professionals, higher supervisory/technicians*Left      -0.010        (0.05)  Lower managers/professionals, higher supervisory/technicians*Right      -0.168**        (0.05)  Intermediate occupations*Left      -0.079        (0.05)  Intermediate occupations*Right      -0.189**        (0.06)  Small employers and self-employed*Left      -0.162*        (0.08)  Small employers and self-employed*Right      -0.208**        (0.08)  Lower supervisors and technicians*Left      -0.139*        (0.05)  Lower supervisors and technicians*Right      -0.160**        (0.06)  Lower sales and service *Left      -0.113*        (0.05)  Lower sales and service *Right      -0.249***        (0.06)  Lower technical*Left        -0.075        (0.05)  Lower technical*Right      -0.390***        (0.06)  Routine*Left      -0.126*        (0.05)  Routine*Right      -0.380***        (0.06)  Intercept 3.165*** 3.263*** 3.128*** 3.116*** 3.164*** 3.217***   (0.03) (0.03) (0.03) (0.04) (.04) (0.05)  Number of individuals 33678 33678 33678 33678 33678 33678  Number of countries 26 26 26 26 26 26  adj. R2 0.152 0.147 0.154 0.154 .174 0.179  Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Note: The “contrasts” column indicates significant differences between class situations for models 1, 3 and 5; between educational categories for models 2, 3 and5; and party affiliation for model 5 (*p < .05).           46  Table 2.3. Average marginal effects for Model 4 Class Less than secondary Secondary/Higher Secondary Above higher secondary University/graduate studies a Large employers, higher managers/professionals 3.914*** b, c, e, f, g, h 3.926*** b, c, d, e, f, g, h 3.893*** b, c, e, f, g, h 3.772*** b, c, e, f, h b Lower managers/professional, higher supervisory/technicians  4.193*** a, c, d 4.024*** a, c, d, e, f, g, h 4.054*** a  3.966*** a c Intermediate 4.153*** a, f 4.129*** a, b, g, h 4.053*** a 4.016*** a d Small employers and self-employed 4.042*** b, c, f, g, h 4.110*** a, b, g, h 4.022***   3.963***  e Lower supervisor and technician 4.218*** a, d 4.181*** a, b, g, h 4.068*** a 3.967*** a f Lower sales and service 4.289*** a, c, d 4.154*** a, b, g, h 4.097*** a 4.051*** a g Lower technical 4.242*** a, d 4.251*** a, b, c, d, e, f 4.166*** a 3.831***  h Routine 4.231*** a, d 4.241*** a, b, c, d, e, f 4.156*** a 4.018*** a *** p < 0.001 Note: Letters next to the average marginal effects indicates the contrasts of predictive margins which are significant for classes across levels of education (*p < .05).    Table 2.4. Average marginal effects for Model 6 Class Left  Centre Right   a Large employers, higher managers/professionals 4.190*** b, c, e, f, g, h  3.958*** c, d, e, f, g, h 3.563*** b, c, d, e, f, g, h   b Lower managers/professional, higher supervisory/technicians  4.259*** a  4.036***  c, e, f, g, h  3.809*** a, c, d, e, f, g, h   c Intermediate 4.264*** a  4.111*** a, b 3.906*** a, f, g, h   d Small employers and self-employed 4.173*** e, g  4.102*** a 3.916*** a, b, g, h   e Lower supervisor and technician 4.278*** a, d  4.184*** a, b 3.950*** a, b, g, h   f Lower sales and service 4.260*** a,  4.141*** a, b 3.995*** a, b, c, g, h   g Lower technical 4.293*** a, d  4.136*** a, b 4.131*** a, b, c, d, e, f   h Routine 4.253*** a  4.147*** a, b 4.133*** a, b, c, d, e, f   *** p < 0.001 Note: Letters next to the average marginal effects indicates the contrasts of predictive margins which are significant for classes across the political spectrum (*p < .05).   47  Figure 2.2. Linear predictions of viewing inequality critically by class and party affiliation  48  2.9. Discussion and Conclusion  Critics of class analysis contend that class differences in life chances and political attitudes have diminished. The weight of the evidence from this analysis not only casts considerable doubt on those claiming the demise of class (Beck and Beck-Gernsheim 2002; Giddens 1991; Inglehart and Rabier 1986; Kingston 2000; Pakulski and Waters 1996), it provides evidence in support of the separation of social class and status in terms of perceptions of inequality, and also resonates with research on the impact of vote choice in shaping orientations towards economic and political equality (Andersen and Heath 2002; Caínzos and Voces 2010; Evans 2002; Sosnaud et al. 2013; van der Waal et al. 2007).  2.9.1. Class, Education and Perceptions of Inequality  I have shown that, in the aggregate, social class influences how people perceive income inequality in their society. More specifically, when compared to the most advantaged class, the other classes encompassed by a service relationship are less critical of income inequalities than those classes under a labour contract. A plausible reason as to why this is the case is that critical views towards inequality are more strongly held by those who are most affected by it. Large employers and higher professionals are mostly sheltered from the negative impacts that material inequality brings and so are less likely to perceive inequality let alone feel strongly against it. Indeed, they may benefit greatly from rising inequality in terms of their material welfare. Their counterparts in the routine and lower technical classes, on the other hand, are more likely to be exposed to the negative impacts of inequality and so are more cognizant and critical of it. These results showcase the class dissensus with regards to inequality views.  While education has been identified as an important influence in research on subjective inequality, it has often been argued those with higher levels of education will be more critical of inequalities (McCall 2013). My findings show that those with limited education tend to be more critical of income inequality. This goes against the theory that those with more education tend to be more liberal in 49  their views towards inequality (McCall 2013). I find no evidence that those who are more highly educated are more critical of inequality in their society. Instead, these new results point towards the purchase of the ideological refinement perspective (Jackman and Muha 1984).  From this perspective, individuals who progress deeper into the education system may internalize the authority granted to them and thus may embody individualistic attitudes towards society. Those who are highly educated have in their possession the skills needed to “protect their interests in a more complex society” (Jackman and Muha 1984: 761). These more individualistic attitudes may divert value preferences away from more equitable measures and promote values less conducive to viewing inequality critically. From this perspective, it is not that surprising that individuals possessing higher levels of education are not opposed to inequality in comparison to their counterparts with lower levels of education.  Further, under our neoliberal times, the concept of progressive universities has been challenged as higher education has been re-designed across the world (Davies and Bansel 2007). Increasing tuition fees, an emphasis on employability, and a general implementation of market values in higher education has been a general trend in many universities across the globe (Baltodano 2012). At the same time, the private benefits of tertiary education are often used as reasons to justify increased tuition fees and the reduction of public funding. Within higher education for instance, critical intellectual debate is often being replaced with performativity, audits, strategic planning, and quality assurance measures (Olssen and Peters 2005). The relationship between class and education in predicting inequality views are intriguing.  Although education partially reflects social class advantages, the present study indicates that they are in fact conceptually and empirically distinct forms of stratification. The relationship between lower class positioning and disapproval of inequality largely endures as education and various other controls are included in the models which contradicts the claims made by anti-class theorist who suggest that competing forms of self-interest have undermined the impact of class. While members of the disadvantaged classes typically possess lower levels of education than their counterparts in the more 50  advantaged classes, it is not just their limited education which leads them to be more critical of inequalities; the characteristics of the class structure also shape their opinions towards inequality. Class, however, is not equally salient across educational levels. While it predicts differences in attitudes to inequality among the relatively less educated, those with above higher secondary education do not differ in their views towards inequality along class lines. The declining significance of class may apply to those who are more educated, perhaps a postmaterialist thesis is restricted to those who are highly educated. Taken together, these findings suggest that future research on attitudes towards inequality can be improved by jointly modelling the effects of both social class and education as an indicator of status. If the study of subjective inequality is undertaken using some one-dimensional measure of class or by conflating class and status, a weaker appreciation of how both structure perceptions of inequality would have resulted. Put differently, status models should complement, rather than replace models based on the social relations of production.  I must be clear that I cannot be sure that social class and education causally influence critical inequality views. What I demonstrate is the possible relationship between class, education and critical inequality views while controlling for additional factors which may also be important in determining these views. Post-treatment bias with respect to the class coefficients could also be an issue. Education precedes employment and is a main device for developing human capital and filtering individuals into specific occupations. As the composite measurement of class includes an occupation component and education is typically causally prior to occupation, it is possible that education could be confounding the occupational component of the class measurement. In other words, education is antecedent to occupation and thus the component of the class measurement attributed to occupation may be accounted for by educational attributes. As a result, by being partly comprised of a component related to education, the estimates generated from class could be biased.  2.9.2. Class Politics and opposition to Inequality 51  This study supports previous research which shows that values associated with vote choice influence opinions towards economic inequality (Andersen and Heath 2002; Caínzos and Voces 2010; Evans 2002; Sosnaud et al. 2013; van der Waal et al. 2007). The findings reveal that those who affiliate themselves with the left of the political spectrum are more likely to view economic inequality as unjust in comparison to those who lean right. While vote choice has been recognized as an important influence in shaping people’s perceptions towards inequality, it has not been recognized as important in moderating the relationship between class position and perceptions of inequality. I have shown that the class position a person occupies allows one to predict the degree to which they are critical of inequality; however, what is unique with my study is that I have shown that this greatly varies depending on their political persuasion. Where the class differences in inequality views are clear-cut are with those who are conservative. For this group, critical opinions towards inequality clearly approximate the class structure of society. To put it differently, objection towards income inequality is deeply aligned with social class for those individuals who affiliate politically with the right. Income inequality is a political issue which often is thought to fall along a “Left-Right” divide. The left tends to think of equality as being central to fairness, and those who vote left are often sensitive to gross levels of inequalities, particularly as they are connected to class issues. Left-leaners tend to favour redistribution and reducing inequality regardless of their class position and so should be more critical of inequality based on their political persuasion. Lower class liberals are partly liberals because of their class position. For this group, residing in employment contracts characterized by limited security, increased precarity and low wages may have led them to adopt a political perspective that favours social policies which protect them from the deleterious effects of the labour market. As such, among those with a liberal orientation we see very little class differentiation. Conversely, it is with those who vote conservative where we see class differences in the degree of tolerance towards inequality. Conservatives are much more sensitive to the notion of meritocracy or that there are slackers or benefit cheats. Large employers and higher managers who are conservative are more 52  likely to tolerate inequality because the self-interest connected to their position in the class structure and the attitudes associated with their elected political persuasion are closely aligned. Right-wing political policies overwhelmingly benefit those who reside in more advantaged classes; yet they hurt the other classes as we move along the class structure in terms of disadvantage. Working class individuals who vote conservative tend to be voting against their economic self-interest in that right-wing political parties tend to favour tax cuts for the rich, cutting benefits for the poor, and opposing redistribution. Conservative parties do not defend the interests of the working classes.   When pondering inequalities in their societies, for conservatives in disadvantaged classes, it is the self-interests associated with their class positions which largely determine their views towards inequality. The working classes have a real material stake in reducing income inequality. As such, self-interest connected to the class structure really comes to play an important role in determining preferences for inequality in those who vote right. Especially among conservatives, the working classes do see their interests. Future research in this area should disentangle the puzzle of conservative support among working class people who disapprove of inequality. Perhaps, it is that this group sees other elements of conservative platforms as anti-inequality (e.g. antiglobalization among populist politicians), or that other elements of the conservative ideology are more salient, which effectively overrides economic interests.   53  Chapter 3: Class, Partisanship and The Great Recession: Conflicting Influences on Attitudes Towards Inequality During Economic Crises 3.1 Introduction  In 2008, the world faced a global financial crisis which left millions of people unemployed and many struggling with economic insecurity (Alvaredo, Atkinson, Piketty et al. 2013; Anthony 2008; Atkinson, Piketty and Saez 2011; Galbraith 2014; Luttig 2013; Piketty 2014; Steele 2015). Average incomes also stagnated or fell in most countries post-crisis, from the years 2007 through to 2011/2012 (OECD 2014; OECD 2016). In the United States, Thompson and Smeeding (2011:11) find that “the Gini Index and the P90/P50 and P90/P10 ratios all increased substantially between 2007 and 2009” (see also Smeeding 2012; Thompson and Smeeding 2015). In response to these events, the global community witnessed a surge in public backlash against the capitalist system which culminated in the Occupy Wall Street movement; a movement which pushed the issue of economic instability and income disparities into mainstream discussion.   Economic instability also weighs heavily on non-economic life. Research has found connections between instability and various social issues including health (German and Latkin 2012), increasing rates of criminal violence (UNODC 2012), and happiness (O'Connor 2017). What this growing body of literature reinforces is that periods characterized by economic instability also have important consequences for societies beyond economic issues. Yet, while we know much about the substance of growing instability and are increasingly aware of its negative consequences for a variety of important social indicators, we know relatively little about how and why citizens’ evaluations of their society change in tandem. For instance, although attitudes towards inequality amidst recession has garnered a fair amount of attention, not as much has been done to address the issue of changes among class-groups (Bartels 2013; Brooks and Manza 2013; Kenworthy and Owens 2011; Kenworthy and Owens 2012). Despite the importance of the opinions citizens hold about the state of inequality in their society, we know relatively little about the forces that shape economic opinions during times of crisis. In particular, it is 54  unclear whether individuals positioned in different classes are more or less critical of inequality during periods of economic instability.   A number of foundational theories suggest that critical views towards inequality should increase during periods of economic turmoil (Durkheim [1893] 1984; Meltzer and Richard 1981; Weber [1922] 1978). In contrast, empirical work on the effects of economic recessions on public opinion towards inequality shows mixed results (Brooks and Manza 2013; Kenworthy and Owens 2011; Kenworthy and Owens 2012). Some scholars suggest that economic shocks are shared events that are experienced uniformly by citizens so they have similar impacts on how people from different groups perceive inequality (Cavaillé and Trump 2015; Kenworthy and McCall 2008; Trump 2017). This “parallel publics” hypothesis represents a common thread in the literature examining public opinion responses to economic recessions (Gonthier 2017). For instance, Kenworthy and Owens (2012) found that there was little impact of the Great Recession (TGR) on the share of respondents saying that American society is divided into the haves and the have nots. Other scholars argue that recessions affect groups differently in how they experience crisis, and that individual evaluations of inequality depend on the extent of their exposure to economic insecurity (Hoynes, Miller and Schaller 2012; Kenworthy and McCall 2008; McCall 2013; Moya and Fiske 2017; Rodriguez-Bailon, Bratanova, Willis et al. 2017; Sands 2017). As an illustration, McCall and Manza (2011) looked at critical sentiments towards inequality in the U.S. in 1996, 2000 and 2008 and found that income had a strong effect. Class locations partly determine an individual’s economic vulnerability, which suggests that different social classes experience different likelihoods of negative effects during periods of economic uncertainty (Weatherford 1978; Weatherford 1982). Recessions may have varying impact on people’s perceptions of inequality across different classes. When the economy is in turmoil the effects of class may be more visible and this may play a greater role in shaping individual views towards inequality. Conversely, when the economy is stable, status may be more important. In the absence of strong class cues from the 55  economy during periods of stability, education, a key dimension of status in contemporary societies, may offer a source of variation in responses towards inequality. Animating this chapter is the contention that critical views towards inequality during recessions are shaped by class but also by politics. The claim that citizen’s opinions towards inequality shift in the same direction during times of crisis also needs to be reconsidered given the important role that politics plays in shaping these views. As an alternative to class, some analysts have shown the explanatory power of party affiliation in explaining perceptions of inequality (De Vries et al. 2018; Evans and Andersen 2006; Evans and Pickup 2010; Pickup and Evans 2013). Indeed, in the previous chapter I showed that the relationship between class and views towards inequality varies depending on political persuasion. This makes sense because dissatisfaction with income inequality is typically what we think of as a political left-right issue. For example, conservatives tend to be less critical of inequality because this is viewed as being associated with individual rights, abilities and responsibilities. In fact, as a response to threats to the social order, people often flock to conservative ideologies, which suggests a possible connection between economic threats and support for conservative ideas (Payne 2017). In this way, disapproval towards inequality at particular times may relate to class but may also be connected to politics (Singer 2011; van der Brug et al. 2007; Weatherford 1982).  Contemporary accounts of peoples’ perceptions of inequality, however, often draw on insights from theories related to class or from political affiliation, but do not consider the interplay between these. During crises it is possible that perceptions of inequality are disparate across classes and are not powerfully affected by political conditioning. In other words, when the economy is in turmoil, the effects of class are more visible and party affiliation may play a reduced role in explaining why individuals view inequality critically. On the other hand, when the economy is stable it is possible that the “partisan contamination” of individual views towards inequality is more likely (Evans and Andersen 2006: 195). The relationship between class and partisanship on viewing inequality critically may be different during recessions as opposed to when the economy is relatively stable. In particular, party affiliation may reduce 56  the strength of the relationship between class and perceptions of inequality, but this may not happen when economic conditions are bad. In this chapter, I blend these aspects into a comprehensive account of people’s views towards inequality by positing that partisanship interacts with class in shaping perceptions of inequality and explore how this relationship is affected by an economic shock. What motivates individuals to think critically about inequality following a major economic shock? Are critical perceptions of inequality disparate across class and party and how are these shaped by TGR? In this chapter, I leverage data collected before and during TGR in order to explore whether individuals from different social classes and who might have been differentially affected by this event, respond similarly or differently to an economic shock. Here, I offer a more well-rounded approach by exploring whether the relationship between class, politics and inequality views is different when there is a shock to the economy as opposed to when it is relatively stable.  I endeavour to improve on previous studies in three ways. First, exploring inequality views by analyzing an extensive dataset of more than 20 countries over 17 years, which is broader in scope than similar research (McCall and Manza 2011; Svallfors 2006). Second, by examining whether dissatisfaction with inequality has become more pronounced along class lines after TGR, which has yet to be explored in previous research. Lastly, by disentangling how the potential class-polarization in inequality views is affected by party affiliation and whether this changes during an economic recession.  3.2 Public Opinion Shifts to Economic Shocks  The economic downturn in 2008 appears to have revitalised concern about the concentration of income and wealth at the top and subsequent dearth of it at the bottom. Recessions bring poverty and inequality into the limelight. They are also periods characterized by increased economic insecurity (Brooks and Manza 2013). Economic insecurity is not, however, distributed evenly across the population; it is concentrated in the more disadvantaged occupations. Recessions thus provide a fruitful site with which to examine how economic fluctuations may reinvigorate dormant class alignments. Changing 57  economic conditions also provide a context in which to examine shifting perceptions of what are regarded as fair or unfair levels of inequality. So, too do they allow one to explore classic sociological theories which suggest that the implications of social class on individual consciousnesses are more pronounced when the class structure and its inherent societal divisions may be more visible.  The idea that critical sentiments may rise in accordance with economic instability has been a long-standing theory in sociology. For instance, in his work on suicide, Durkheim ([1897] 2002) noted that economic fluctuations may affect individuals’ sense that the world is stable and predictable. In the Division of Labour in Society, he also argues that sudden changes in the economy can lead to anxiety and a fear for the future as individuals lose a sense of control over their social standing (Durkheim [1893] 1984:310-16). Durkheim ([1893] 1984: 314) goes on to suggest that inequality is tolerated more in “traditional” societies because people grow accustomed to it as it is regarded as natural. When organic solidarity dominates, inequalities are not tolerated, and if they are allowed to grow, serious consequences may follow: “At the very moment when the flood tide grows more violent, the dyke that contained it is breached. Thus the situation becomes much more dangerous . . . If societies attempt—and they should attempt—to eliminate external inequalities as much as possible, it is not only because the undertaking is a noble one, but because in solving this problem their very existence is at stake. . .  Equality in the external conditions of the struggle is not only needed to secure each individual to his function, but also to link these functions with one another.” (Durkheim [1893] 1984:315-16)  Weber also addresses the issue of rising discontent during periods of economic turmoil. During periods of relative market stability society tends to be differentiated by status, which for Weber ([1922] 1978:38), is connected to the “social estimation of honor” attached to certain positions; while class, which is grounded in the social relations of economic life, is placed on the backburner. During periods of 58  economic turmoil, the situation is reversed, with class pushed to the fore as stratification based upon status recedes: “When the bases of acquisition and the distribution of goods are relatively stable, stratification by status is favoured . . . economic transformation threatens stratification by status and pushes class situation into the foreground. Epochs and countries in which the naked class situation is of predominant significance are regularly periods of technical and economic transformations. And every slowing down of the change in economic stratification leads, in due course, to the growth of status structures and makes for a resuscitation of the important role of social honor” (Weber [1922] 1978:38)  Does economic instability have implications for awareness of inequalities? The consequences of economic instability are apparent and one central question has been how individuals from various stratums in society will react. A now classic model developed by Meltzer and Richard (1981) suggests public discontent, including the formation of social movements and the potential for revolutions. This notion is connected to relative deprivation thinking, which assumes that when an economy is in crisis, or when there is a downturn, people become more aware that they are worse off than they were in the past (Brooks and Manza 2013; Jetten, Mols, Healy et al. 2017; Mols and Jetten 2017). With instability comes visible manifestations of inequality such as poverty, social conflict and economic precarity. Accordingly, economic instability could lead to increased awareness and concern including more critical attitudes towards income inequality.  3.3 Opposition and Contentment to Economic Instability  Classic theory in sociology and political science suggests that periods characterized by economic instability should be matched with increasing critical sentiments towards inequality (Durkheim [1893] 1984; Meltzer and Richard 1981; Weber [1922] 1978). Despite these theories, much of the literature in 59  the social sciences takes a skeptical view (Bartels 2013; Brooks and Manza 2013; Cavaillé and Trump 2015; Durante and Fiske 2017; Jetten et al. 2017; Laurin, Gaucher and Kay 2013; Moya and Fiske 2017; Rodriguez-Bailon et al. 2017; Sachweh ; Trump 2017). Investigations into individual responses to economic downturns show mixed results, with some suggesting that individuals respond critically to economic shocks (Curtis and Andersen 2015; Osberg and Smeeding 2006; Piketty 2000); while others contend that most show little interest or awareness of the broader economic context (Kenworthy and McCall 2008; Margalit 2013; McCall 2013). As an illustration, McCall (2016), found that concerns about inequality stabilize or rise during the initial years of recovery from a recession. Indeed, she found the peaks of concern about inequality are connected to the perceived negative consequences of inequality rather than with perceptions of the level of inequality itself.  The lack of opposition to income inequality during periods of economic instability is intriguing considering the classic theory about public opinions suggests that economic self-interest is a key determinant of demand for redistribution (Meltzer and Richard 1981). Research in the area of opinions towards inequality is, however, restricted because of the focus on aggregate patterns, which may obscure important group-specific trends. Hardships may trigger very different reactions across classes. For instance, those residing in social positions with differing degrees of advantage in terms of economic security could respond to economic change in antagonistic ways while country-wide patterns may appear relatively stable. Focusing on aggregate patterns also makes it impossible to discern whether any overall rising trend in disapproval reflects changes in particular groups or a more general trend.  What is lacking with current research on public opinion shifts to economic recessions is that these vary because people may be sheltered from the negative effects of recessions based on their social position (Andersen and Yaish 2012; Kenworthy and McCall 2008; McCall and Percheski 2010; McCall and Manza 2011). In particular, a limitation of existing research in this area is the lack of consideration given to the different degrees of severity of the recessions impact associated with social class which may affect how individuals interpret the meaning of inequality. Building on earlier research, this chapter 60  examines the relationship between class and public opinion shifts towards economic shocks (Etzioni 2014; Kenworthy and Owens 2011; McCall and Percheski 2010; McCall and Manza 2011; Walasek and Brown 2016). 3.4 Theorizing Class Responses to Economic Crises An individual’s economic vulnerability during crises differs depending on their location in the class structure (Hout, Levanon and Burak 2011; Hoynes et al. 2012). For example, TGR may have affected workers in manual/non-standard jobs more strongly because the compensation schemes of the working classes provide less job security, offer reduced benefits, and their wages are more susceptible to decline during recessions (Rose and Harrison 2007; Rose and Harrison 2010). The working classes, in particular, should react to economic instability by holding more critical sentiments towards material inequalities evident in their society (Dodson 2017). In comparison, the upper classes, enjoy compensation agreements which offer greater job security (e.g. protection against redundancy, length of notice required to terminate contracts), such that members may be less negatively affected by periods of economic instability and hence may react less critically. What these scenarios illustrate is that the more disadvantaged classes could be regarded as a volatile segment of society, which may erupt in times of economic crisis; whereas, the advantaged classes are a more tranquil group and are more tolerant of inequalities even during periods of economic turmoil.  Despite the class-based factors which may lead individuals to react differently to TGR an argument persists that we are witnessing the dissolution of class in shaping value preferences towards economic issues (Bauman 1982; Beck and Beck-Gernsheim 2002; Clark and Lipset 1991; Giddens 1991; Kingston 2000). The death of class perspective would view this turbulence not as an indication of political antagonism between classes defined in terms of the social relations of production, rather these would be regarded as discord between status groups with conflicting value commitments (Pakulski and Waters 1996). This perspective holds that in Western countries, the class-basis of political representations and related demands are in steady decline.  61  Death of class scholars contend that there has been a progressive demise of collective identities which has ushered in an era where social class no longer impinges on people’s perceptions and attitudes as it did in the past (Bauman 1982; Beck and Beck-Gernsheim 2002; Clark and Lipset 1991; Giddens 1991; Kingston 2000). Cultural issues are thought to prevail over traditional economic issues such as attitudes towards redistribution and critical sentiment towards inequality. The argument for the diminishing relevance of class in contemporary societies contains different factors, but some common themes include: new political cleavages emerging, changes in the labour market and the associated decline in working class occupations; and rising levels of material welfare and affluence. For death of class proponents and their critics, the issue of awareness of inequality is a key marker of class-relevant politics (Edlund and Lindh 2015).  Scholars studying the declining significance of class in contemporary societies have focused around a number of important empirical issues, such as class politics (Evans 2002; Evans and Tilley 2012b), trends in educational and occupational inequalities (Goldthorpe 1996; Weeden and Grusky 2012), and income disparities (Wodtke 2016; Wodtke 2017; Zhou and Wodtke 2018). Although there have been a number of scholars who have challenged this assertion on a theoretical and empirical level (Andersen and Curtis 2012; Atkinson 2010; Haddon 2015; Wodtke 2017), few have provided an empirical assessment of the endurance of class following an economic shock (for an exception see McCall and Kenworthy 2009; McCall and Manza 2011).  Who experiences changes in their critical assessments of inequality following a period of salient economic threat? The principle puzzle addressed here is how the economic shock of TGR shaped classed attitudes towards income inequality. Given the economic crisis which affected most of the Western world, class-related differences in economic interests might be expected to increase rather than diminish. This is because during recessions, the visibility of the different economic interests associated with class are heightened. Such occurrences may enhance the relationship between class and inequality views because the objective structure of the class system is more visible. During periods where the implications of class 62  are evident, such as during a recession, people, with a common class position, may be more likely to share common views towards society. This suggests:  Hypothesis 1: The class differences in viewing inequality critically will be stronger in 2009 than in earlier periods.  3.5 Politics During Crisis  In politics, the left is thought to be more critical of inequality while the right is more receptive (De Vries et al. 2018; Evans and Andersen 2006; Evans and Pickup 2010; Pickup and Evans 2013). Crises may magnify these different partisan responses (Brooks and Manza 2013). During periods of economic turmoil, it is often suspected that individuals would flock to progressive politics (Bartels 2013). In the United States however, the political response to the Great Recession was not a shift to the left as has been foreshadowed, but largely a shift to the right that included the Tea Party movement, right-wing populism advocating reduced government, and bailouts (Bartels 2013). For example, Brooks and Manza (2013) found that in the American context, economic change had little effect on attitudes towards social policies. Rather than gravitate to increased government support, they found that the public demanded reduced government involvement. Indeed, in many countries, progressive politics failed to take hold in the wake of the Great Recession (Margalit 2013; Milojev, Greaves, Osborne et al. 2015).  Those who are committed politically to the left, tend to view economic inequality as unjust and maintain that income inequalities are too large (Evans and Tilley 2017). Conversely, conservatives often underestimate the degree of economic inequality in their society (Norton and Ariely 2011). As an illustration, McCall (2013) finds that in the United States, Democrats are more inclined to oppose current levels of inequality than Republicans, and liberals are more likely to oppose current levels of inequality than conservatives. Following established research, it is plausible to assume that left-wing party supporters are more perceptive of income inequality during periods of economic instability, in comparison to those who favour right-wing political parties (Evans and Andersen 2006; Evans and Tilley 63  2017; Norton and Ariely 2011). As such, I suspect that during periods of economic instability this relationship should persist and become more evident, in that the relationship between left-wing party support and critical views towards inequality should be strongest in 2009, compared to the earlier periods. Hypothesis 2 tests this assertion:  Hypothesis 2: The positive relationship between left-wing party affiliation and critical perceptions of inequality will be stronger in 2009 than in earlier periods.  3.6 Class Politics During Crises  One of the tenets of the death of class perspective is that since the 1970s, “the significance of class as a basis for political identification and behavior and as a force for change has been declining” (Pakulski and Waters 1996: 132). One way recent research has confronted this assertion is by showing the continued impact that class has on political orientations (Evans and Tilley 2012a; Evans and Tilley 2012b; Evans and Tilley 2017). Indeed, in the previous chapter, I demonstrated the interplay between class and party affiliation in predicting opinions towards inequality.  The relationship between class, party affiliation and critical views towards inequality may also depend on the state of the economy. For example, research has shown that the class-basis of voting is more evident when government liabilities in the economy are more clearly identified, such as during economic recessions (Giuliani and Massari 2017). Class differences in politics may be in response to changing economic conditions, in that individuals tend to vote along class lines when related economic issues are paramount (Duch and Sagarzazu 2014). The greater liability of the working-class during recessions to the cost of unemployment or precarity may give this class an interest in policies characteristic of left-wing parties. Implicit in this characterization is the notion of considerable heterogeneity in how certain class groups vote, in how these groups experience economic shocks, and finally how these together shape individual concerns towards inequality. 64  Shocks to the economic system may cut across party lines. When the economy is more volatile, the moderating role that partisanship plays in the relationship between class and inequality views may weaken. When it is stable, the moderating role that partisanship plays in the relationship between class and inequality views may strengthen. In other words, when economic conditions are bad, the effects of class are more visible and the capacity of party affiliation to reduce the impact of class on inequality views may decrease. In contrast, when conditions are good, the class structure is less visible, meaning the ability of party affiliation to reduce the impact of class on inequality views may increase. Based on this consideration, I propose a three-way interaction between class, party affiliation and an economic shock and test the following hypothesis: Hypothesis 3: Following an economic shock, the class differences in inequality views vary less by party affiliation. 3.7 Methods 3.7.1. Data Analyses draw on data from the International Social Survey Programme (ISSP). Of particular interest is the survey module on Social Inequality which was carried out in 1992, 1999 and 2009. Data is missing in some countries for the various years. Data for 1992 is available for 18 countries (countries that did not participate are Chile, Cyprus, France, Japan, Latvia, Portugal, Spain and Switzerland). Data for 1999 is available for 24 countries (Italy and Switzerland did not participate). For 2009, data is available for all countries except Canada, which is not available due to low response rates. The lower number of countries included in 1992 means that my ability to ascertain the generality of patterns cross-nationally is less in this year compared to 1999 and 2009. Nevertheless, one benefit of analyzing the data set is that the 1992-2009 time-frame enables me to test the hypotheses with data collected before and after the 2008 financial crisis. The three time points also allow me explore change and stability in attitudes about inequality.  65  A noted limitation of this analysis is that the effects of TGR on inequality views are not completely identified by the inclusion of data from the survey year 2009. That is, other factors that occurred in various countries leading up to TGR could be affecting public opinions towards inequality. For instance, a number of global events happened in the decade long gap between 1999 and 2009, which may mean that these could have had implications on inequality views more so than TGR. What this means is that including the survey year 2009 may not fully capture the effects of TGR, other factors could be impacting the relationship between class, partisanship and inequality views.  3.7.2. Dependent Variable  To gauge public awareness of income inequality, I use the following question: “How much do you agree or disagree with the statement ‘Differences in income in [respondent’s country] are too large?’”. There are five possible responses: strongly agree, agree, neither agree nor disagree, disagree, and strongly disagree. As in the previous chapter, I treat the variable as an interval level measurement such that higher values equate with more critical views towards inequality.3 3.7.3. Key Independent Variables  Class is operationalized using a variation of the Goldthorpe (EGP) class schema known as the European Socioeconomic Classification (ESeC) (Rose and Harrison 2007; Rose and Harrison 2010). The ESeC, which is inspired by a Weberian interpretation of class, is a newer version of Goldthorpe’s schema. As those in similar class positions possess certain resources by virtue of the positions they occupy, it is understood that they will share comparable possibilities and restrictions in relation to “life-chances”. Arguably, Goldthorpe’s schema is now the most influential categorical measure of social class and is often regarded as the most valuable way of measuring, empirically, class-based inequalities (Evans 1996; Haddon 2015; Marks 2005; Marshall, Newby, Rose et al. 1988; Svallfors 2005). The schema contains                                                      3 I also treated the variable as ordinal and ran ordered logistic regression models. The results for the overall relationships between class, partisanship and inequality views over time were similar, besides the following: the interaction between intermediate class*1992 is not significant; the interaction between left-wing part support*1992 is not significant; and the three-way interaction between working class*left*1999 becomes significant. Results for this modelling strategy are provided in Table B.1 in the Appendix. 66  both construct and criterion related validity and provides a useful list of class situations when one wishes to assess cross-national differences (Bergman and Joye 2005; Evans 1992; Evans and Mills 1998; Evans and Mills 2000).  The aim in taking a class perspective is to research how class situations create different sets of attitudes for the individuals who occupy them and to explore the class-linked variation in terms of how individuals perceive inequality. To reside in a class location means that one is subjected to a set of mechanisms that impinge on the lives of individuals as they interpret and navigate the social world. While these situations can determine the material interests of the individuals or the resources necessary to pursue these interests; my main aim is on elaborating how these situations shape their views towards inequality. In this chapter and the next, I elect to use the three-class version. Hence, as illustrated in Table. 3.1, this includes the salariat, intermediate and working class. I use the three-class version so as to enable an easier interpretation of the results of the interactions between class, party affiliation and survey year. Collapsing the eight-class model into a three-class version retains the basic divisions between the three forms of employment regulations (service relationship, labour contract, and mixed combinations). Table 3.2 provides the class compositions for the various countries and years. In the statistical models, the salariat class is the reference group.  67  Table 3.1. Collapsing the eight-class model into a three-class version Class 8 class version 3 class version Large employers, higher grade professional, administrative and managerial occupations 1 1+2 (Salariat Class)  Lower grade professional, administrative and managerial occupations and higher-grade technician and supervisory occupations 2 Intermediate occupations 3 3+4+5 (Intermediate Class)  Small employers and self-employed 4 Lower supervisory and technicians 5 Lower services, sales and clerical occupations   6 6+7+8 (Working Class)  Lower technical occupations 7 Routine occupations 8    68  Table 3.2 Class compositions for each country and year (proportions) Country Year N Salariat Class Intermediate Class Working Class Australia 1992 1,785 .44 .28 .28  1999 1,466 .38 .25 .37  2009 1,390 .49 .28 .23 Austria 1992 478 .42 .20 .37  1999 498 .28 .39 .33  2009 886 .29 .38 .33 Bulgaria 1992 878 .33 .11 .56  1999 1,004 .24 .17 .59  2009 806 .33 .14 .53 Canada 1992 697 .52 .24 .24  1999 797 .49 .27 .24  2009 - - - - Chile 1992 - - - -  1999 684 .19 .35 .47  2009 1,296 .17 .33 .50 Cyprus 1992 - - - -  1999 800 .28 .29 .43  2009 834 .30 .34 .36 Czech Republic 1992 654 .23 .28 .49  1999 1,754 .26 .24 .50  2009 1,064 .27 .20 .53 France 1992 - - - -  1999 1,201 .58 .27 .14  2009 2,494 .41 .31 .27 Germany 1992 2,724 .27 .28 .45  1999 749 .36 .33 .30  2009 1,278 .34 .27 .39 Hungary 1992 1,110 .22 .19 .59  1999 1,078 .24 .21 .55  2009 927 .18 .19 .63 Italy 1992 522 .43 .25 .31  1999 - - - -  2009 859 .41 .32 .27 Japan 1992 - - - -  1999 724 .13 .51 .36  2009 716 .26 .38 .36 Latvia 1992 - - - -  1999 696 .38 .19 .43  2009 969 .28 .20 .52 New Zealand 1992 759 .43 .31 .26  1999 779 .42 .24 .34  2009 622 .48 .37 .15 Norway 1992 1,328 .39 .21 .40  1999 1,153 .40 .27 .33  2009 1,371 .43 .32 .24 Philippines 1992 689 .09 .21 .11  1999 633 .61 .49 .48 69   2009 1,009 .29 .30 .41 Poland 1992 1,521 .19 .35 .45  1999 502 .28 .28 .44  2009 1,105 .34 .25 .41 Portugal 1992 - - - -  1999 942 .17 .29 .54  2009 868 .22 .29 .49 Russia 1992 1,295 .45 .13 .42  1999 706 .35 .21 .45  2009 1,395 .34 .16 .50 Slovak Republic 1992 391 .25 .21 .54  1999 878 .25 .22 .52  2009 996 .22 .26 .53 Slovenia 1992 650 .23 .29 .48  1999 895 .30 .22 .48  2009 861 .32 .27 .41 Spain 1992 - - - -  1999 902 .14 .27 .59  2009 855 .17 .25 .58 Sweden 1992 674 .37 .09 .55  1999 1,040 .34 .26 .40  2009 1,073 .41 .24 .35 Switzerland 1992 - - - -  1999 - - - -  2009 1,115 .40 .33 .26 UK 1992 1,050 .33 .24 .43  1999 787 .29 .27 .44  2009 936 .35 .27 .38 US 1992 1,182 .35 .24 .41  1999 1,201 .36 .20 .44  2009 1,512 .38 .25 .37  70   Party affiliation is included as a categorical variable including: “Left”, “Centre”, and “Right” (the reference group). Proportions for the key variables by year and country are provided in Table 3.3.4  All models control for status, gender, age, marital status, and religion. Status is operationalized through education, which is measured categorically by qualification obtained. The categories consist of “Less than secondary”, “Secondary/higher secondary”, “Above higher secondary” and “University degree” as the referent category. I include age as a control as it has been argued that younger people look more critically at income inequality and recessions may have cohort effects, in that they may exert their effects among those who are in their formative years during recessions (Hadler 2005; Kenworthy and Owens 2011). Inequality has been shown to be legitimated by various religions, as such all models control for religion using Roman Catholic as the reference category (Wisman and Smith 2011).                                                      4 Models were also run including those missing on the party support variable (i.e. non-voters and those not classified) and the results are provided in table B2 in the Appendix. In model 2, omitting those who are not politically engaged biases the estimates for the classes and survey year 1992. In model 3, the interaction between working class and 1992 is biased. For model 4, the estimates for 1992, all classes, the interactions between left and centre in 1999 are biased. Finally, in model 5, the estimates for 1992, and left*1999/centre*1999 are biased. As with the previous chapter, the results generated in this chapter only pertain to those who are politically engaged.  71  Table 3.3 Proportions for key variables Country Year Mean Subjective Inequality  Proportion voting right Proportion voting left Australia 1992 3.62 .46 .49  1999 3.76 .48 .46  2009 3.91 .46 .52 Austria 1992 4.09 .12 .49  1999 4.23 .46 .52  2009 4.30 .49 .51 Bulgaria 1992 4.78 .25 .42  1999 4.78 .49 .44  2009 4.5 .35 .45 Canada 1992 3.81 - -  1999 3.76 .12 .10  2009 - - - Chile 1992 - - -  1999 4.30 .25 .44  2009 4.19 - - Cyprus 1992 - - -  1999 3.65 .45 .32  2009 3.73 - - Czech Republic 1992 3.90 .40 .14  1999 4.40 .42 .45  2009 4.28 .32 .55 France 1992 - - -  1999 4.40 .21 .48  2009 4.56 .23 .42 Germany 1992 4.23 .33 .58  1999 4.04 .33 .59  2009 4.37 .29 .64 Hungary 1992 4.20 - -  1999 4.58 .43 .51  2009 4.74 - - Italy 1992 4.38 - -  1999 - - -  2009 4.59 .32 .54 Japan 1992 - - -  1999 3.90 .61 .16  2009 4.11 .44 .06 Latvia 1992 - - -  1999 4.52 - -  2009 4.52 .16 .47 New Zealand 1992 3.89 - -  1999 3.88 .64 .36  2009 3.69 - - Norway 1992 3.75 .30 .47  1999 3.79 .37 .42  2009 3.52 .44 .40 Philippines 1992 3.49 - - 72   1999 3.67 .34 .09  2009 3.28 .27 .14 Poland 1992 4.18 - -  1999 4.32 .28 .60  2009 4.36 - - Portugal 1992 - - -  1999 4.75 .04 .61  2009 4.54 .06 .62 Russia 1992 4.37 .13 .35  1999 4.71 .27 .55  2009 4.54 .38 .38 Slovak Republic 1992 4.29 .19 .21  1999 4.65 .20 .28  2009 4.52 .13 .56 Slovenia 1992 4.26 .22 .21  1999 4.36 .43 .14  2009 4.51 .18 .27 Spain 1992 - - -  1999 4.22 .25 .47  2009 4.19 .18 .53 Sweden 1992 3.60 - -  1999 3.87 - -  2009 3.93 .32 .51 Switzerland 1992 - - -  1999 - - -  2009 4.20 .51 .42 UK 1992 4.08 .44 .40  1999 4.05 .34 .50  2009 3.97 .44 .40 US 1992 3.92 .29 .15  1999 3.76 .25 .34  2009 3.76 .24 .36 73  3.8 Analytical Strategy In this chapter, I employ linear fixed effects regression models. All models account for the time invariant characteristics from the independent variables, which enables me to assess the net affect of each. The survey wave variables are included as a series of dummy variables as I am interested in the specific time covariates. Further, as between-country differences are controlled for in the models with country fixed effects, all models offer a fairly strict test of my hypotheses about possible changes in the salience of class. When holding the country-level constant in order to isolate across country effects, I can determine the extent to which the relationship between class and subjective inequality varies temporally. At this point, I do not introduce explanatory variables at the country-level as I am predominantly interested in employing fixed effects models for descriptive purposes to explore trends in class polarization following an economic shock. Further, to maintain comparability across models, I drop observations with missing values on any variable used in any of the models. As in the previous chapter, all models apply probability weights. I begin by fitting models that predict subjective inequality from survey year and controls only. Model 1 in Table 3.4 includes the year dummy variables alongside the controls which reflect how each year impacts subjective assessments of inequality. Here, I include 2009 as the reference year. The coefficients of interest are thus in comparison to a period shortly after an economic shock. Significant effects for the year dummy variables would indicate that inequality views changed in the given year in comparison to the base year of 2009. In model 2, I include class and party affiliation alongside survey year to explore whether time trends in attitudes to inequality are tied to changes in the class and political composition in the ISSP countries. The task here is to explore relationships between class, party affiliation and inequality views over time.  In model 3, I explore the relationship between class and inequality views over time by including an interaction effect between social class and survey year. Including an interaction term allows me to uncover whether the relationship between class and critical inequality views is positive when the time-74  trend is isolated. This model offers a test for hypothesis 1, as to whether the class differences in viewing inequality critically will be stronger in 2009 than in earlier periods. Model 4 includes an interaction between survey year and party affiliation to test hypothesis 2 that the positive relationship between left-wing party affiliation and critical perceptions of inequality will be stronger in 2009 than in earlier periods. Model 5 includes a three-way interaction between class, party affiliation and year in order to test hypothesis 3 that after a recession, the class differences in critical inequality views vary less by party affiliation.  3.9 Results  To begin, I explore trends in critical views towards inequality. Figure 3.1 provides an illustration of the trends in critical sentiments towards inequality from 1992 to 2009. Here we see critical inequality views increased from 1992 to 1999 and then plateaued. The tendency to view inequality critically reaches its peak in 2009. Figure 3.2 reveals whether the tendency of rising criticism towards inequality varies by social class positioning. We see clear differences in attitudes towards inequality by class. The working class are more critical of inequalities, over time, than the intermediate class, who are more critical than the salariat class. However, the time trends are similar across classes, revealing increases in disapproval of inequality across the class spectrum. For the working class, the mean subjective inequality score was 4.2 in 1992, and this increased to 4.33 in 2009. At the opposite end of the class spectrum, the mean subjective inequality score for the salariat was 3.86 in 1992, increasing to 4.11 in 2009. These findings indicate that social class may be related to critical views towards inequality. However, the findings are only descriptive in nature and need to be shown to hold under the scrutiny of a more complex statistical analysis.75           4.054.074.094.114.134.154.174.194.211992 1999 2009Figure 3.1 Mean inequality views by survey year 76      3.83.944.14.24.31992 1999 2009Salariat Class Intermediate Class Working ClassFigure 3.2 Mean inequality views by social class over time 77  In Table 3.4, I begin to parse out the temporal relationship between class, party affiliation and inequality views. The first column provides the results of Model 1, which explores whether critical awareness of inequality is becoming stronger alongside a period of economic instability. There is no evidence to show that critical views towards inequality were disrupted by TGR. While there is a significant negative coefficient in 1992, which indicates individuals were less critical of inequality in this time period, there is a significant positive coefficient for 1999, which suggests that people were more critical in this time period compared to 2009. More specifically, in comparison to 2009, inequality views were .05 points higher in 1999; whereas they were .03 points lower in 1992. Critical attitudes towards inequality actually decreased slightly after TGR.    In model 2, I include the class, education and party affiliation indicators alongside survey year. The first item to notice is that the coefficients for the survey waves remain statistically significant, indicating aggregate attitudes towards inequality have changed over time net of these aspects of individual social positioning. Inequality views also differed depending on class and party affiliation. In terms of class, the working-class hold critical inequality views .16 points higher in comparison to the salariat class. Those who affiliate with the left hold inequality views which are .34 points higher than those who are right leaning.  In Model 3, I enter interactions between class and year. Year is hypothesised to act as a moderating variable, in that the relationship between class and inequality views should be stronger in 2009 and comparatively weaker in 1999 and 1992. The results of Model 3 present a significant positive coefficients for the interaction of the working class and the intermediate class in 1992 (vs. 2009) in regards to placement on the inequality scale. Thus, the working and intermediate classes were more critical of inequality relative to the salariat in 1992, with attitudes becoming more similar in 2009. The coefficient for 1992 also remains negative and statistically significant (p<.001) which indicates that the salariat had less critical views about inequality in 1992 in comparison to 2009. Jointly the interaction 78  terms are statistically significant (p < .001) which indicates that the relationship between class and perceptions of inequality varies by survey year5. Table 3.5 provides the average marginal effects plus the contrasts of the predictive margins between classes for each survey year. The table reveals that the majority of significant differences between classes occurred in 1992. Between the working and intermediate classes in 1999 and 2009 there are no significant class differences. The salariat significantly differ from the other classes in their inequality views in all survey years. To more clearly illustrate how these class dynamics played out during TGR, in Fig. 3.3, I plot the average marginal effects between class and survey year as regards to inequality views. By looking at the predictions for each class separately I can uncover the class trends in inequality views. The figure shows that inequality views follow a linear trend for all classes from 1992 to 1999, with more critical inequality views in 1999, although there are no significant differences between the working and intermediate classes. The salariat class saw a sharp increase in their disapproval towards inequality from 1992 to 1999; while the working class saw only a slight increase. In response to TGR, all classes were less critical of inequality than they were in 1999. Further, the figure also shows a decrease in the distance between the three classes’ views towards inequality. That is, inequality views come closer together such that the working class differs less from the intermediate and salariat class. While in 1992 the working class held inequality views which were more critical than the salariat and intermediate classes, in 1999 and 2009 the difference between the three class categories is less. Taken together, the findings refute hypothesis 1 that class differences in viewing inequality critically will strengthen over time. Instead of strengthening class differences in perceptions of inequality, economic uncertainty actually muted them. In Model 4, I include an interaction between party support and survey year. There is a significant positive coefficient for the interaction of “left” party support in 1992 (vs. 2009) in terms of placement on                                                      5 Joint tests of significance for all variables across the models are provided in table B.3 in the Appendix. 79  the inequality scale and jointly the interactions terms are significant (p < .001). This suggests, that left leaners were more critical of inequality relative to conservatives in 1992 and attitudes became more similar in 2009. Further, in 1992 the tendency for centrists to view inequality more critically than right-leaners differed significantly from that in 2009. Table 3.6 shows the average marginal effects as well as providing the significance of the contrasts of the predictive margins across the political spectrum for each survey year. The table indicates that there are significant differences between the various party affiliations across the years. From 1992 to 2009, the left is significantly more critical of inequalities than both conservatives and those in the centre.  Figure 3.4 portrays the average marginal effects graphically. Across the political spectrum, inequality views were slightly less critical in 2009 than they were in 1999. Conservatives became more critical of inequality from 1992 to 1999 but are slightly less critical in 2009. Individuals on the left and centre of the political spectrum are fairly consistent in their inequality views over time. As a result, Hypothesis 2, which surmised that the positive relationship between left-wing party affiliation and critical perceptions of inequality will be stronger in 2009 than in earlier periods, is refuted.   In model 5, I test hypothesis 3, which posited that following a recession, the relationship between class and disapproval of inequality will vary less by party affiliation. The results confirm the findings from the previous chapter that class and party affiliation jointly impact attitudes towards inequality (p < .001). However, the joint interactions terms between class, party affiliation and year are not significant. Further, there is also only one significant three-way interaction (p < .05) between the intermediate class centrist voters in 1999. Intermediate class centrists were therefore more critical of inequality than salariat centrists in 1999; after TGR, however, there were no significant difference between centrist intermediates and salariats. In table 3.7, I offer the average marginal effects for the class by party interactions across the years as well as the significance of the predicted margins. If we focus in on those who affiliate with the right, we see that in this group there are more significant between class differences across the years 80  compared to those on the left and centre. The table also reveals that there are more significant differences across partisan lines between the salariat and the other two classes.  Figure 3.5 graphs the average marginal effects of class and party affiliation for each survey year. The first aspect to take notice of is that the impact of class among left wing voters in 2009 was muted as can be seen by the lack of significant differences between the classes. Further, after TGR, the distance between the inequality views of the working class irrespective of party affiliation are similar. For the working classes, inequality views tend to be more similar regardless of politics. Moreover, there is more political variation in condemning inequality among the salariat in 1992 in comparison to the later time periods. What this shows is that the impact that politics has on viewing inequality critically after TGR, as shown by the gap between the predictions, is most evident among the salariat classes.  These results demonstrate that there were greater political differences between the classes in their views towards inequality in 1992, compared to the later time periods. For the working class, there is little political variation across the years, indicating that working class inequality views were less affected by their political perspective from 1992 to 2009. The intermediate class, also show less political fluctuation in 2009, compared to earlier. The salariat has the most political fluctuations in their views towards inequality compared to the other classes, although in this group we also see evidence of fewer political differences in 2009 compared to the earlier years. Taken together, these finding confirm hypothesis 3 which posited that class differences in viewing inequality critically should vary less by party affiliation in 2009.    81  Table 3.4 Results from linear fixed effects regression models and interactions between year, class and political affiliation in inequality views  (Model 1) (Model 2) (Model 3) (Model 4) (Model 5) Contrasts Survey Year       a 1992 (ref: 2009) -0.032* -0.057*** -0.149*** -0.123*** -0.218*** b1, b2  (0.02) (0.02) (0.03) (0.03) (0.05)         b 1999 0.045** 0.037** 0.020 0.064** 0.081* a1, a2  (0.01) (0.01) (0.02) (0.02) (0.04)  Controls       Female 0.124*** 0.115*** 0.115*** 0.115*** 0.111***   (0.01) (0.01) (0.01) (0.01) (0.01)         Age 0.004*** 0.004*** 0.004*** 0.004*** 0.004***   (0.00) (0.00) (0.00) (0.00) (0.00)         married -0.024* -0.002 -0.003 -0.002 -0.002   (0.01) (0.01) (0.01) (0.01) (0.01)         Not religious (ref: Catholic) 0.074*** 0.055*** 0.056*** 0.055*** 0.051**   (0.02) (0.02) (0.02) (0.02) (0.02)         Protestant -0.014 -0.004 -0.005 -0.003 -0.001   (0.02) (0.02) (0.02) (0.02) (0.02)         Other religion 0.069* 0.046 0.046 0.046 0.044   (0.03) (0.03) (0.03) (0.03) (0.03)         Class       c Working Class (ref: Salariat)  0.159*** 0.109*** 0.159*** 0.333*** d2   (0.01) (0.02) (0.01) (0.04)         d Intermediate Class  0.109*** 0.068** 0.110*** 0.193*** c2   (0.01) (0.02) (0.01) (0.04)  Education       e Less than secondary   0.191*** 0.196*** 0.190*** 0.203*** f2, g2 (ref: University/postgraduate)  (0.02) (0.02) (0.02) (0.02)         f Secondary/Higher Secondary  0.158*** 0.167*** 0.158*** 0.171*** e2, g2   (0.02) (0.02) (0.02) (0.02)         g Above higher secondary  0.115*** 0.121*** 0.114*** 0.126*** e2, f2   (0.02) (0.02) (0.02) (0.02)  Party Affiliation       h Left (ref: right)  0.340*** 0.340*** 0.334*** 0.522*** i2   (0.01) (0.01) (0.02) (0.03)         i Centre  0.184*** 0.184*** 0.167*** 0.291*** j2   (0.02) (0.02) (0.03) (0.04)  Two-way Interactions       Working Class*1992   0.165***  0.194**  (ref: salariat, 2009)   (0.03)  (0.06)         Working Class*1999   0.017  -0.049     (0.03)  (0.06)  82         Intermediate Class*1992   0.105**  0.076     (0.04)  (0.06)         Intermediate Class*1999   0.038  -0.040     (0.03)  (0.06)         Left*1992    0.097** 0.099  (ref: Right, 2009)    (0.03) (0.06)         Left*1999    -0.050 -0.084      (0.03) (0.05)         Centre*1992    0.103** 0.128      (0.04) (0.07)         Centre*1999    -0.027 -0.087      (0.03) (0.06)         Working Class*Left     -0.369***  (ref: Salariat Right)     (0.05)         Working Class*Centre     -0.238***       (0.06)         Intermediate Class*Left     -0.213***       (0.05)         Intermediate Class*Centre     -0.159**       (0.06)  Three-way Interactions       Working Class*Left*1992     -0.034  (ref: Salariat Right, 2009)     (0.08)         Working Class*Left*1999     0.098       (0.07)         Working Class*Centre*1992     -0.101       (0.09)         Working Class*Centre*1999     0.039       (0.08)         Intermediate Class*Left*1992     0.047       (0.08)         Intermediate Class*Left*1999     0.052       (0.07)         Intermediate Class*Centre*1992     0.043       (0.10)         Intermediate Class*Centre*1999     0.198*       (0.08)         Intercept 3.456*** 3.120*** 3.148*** 3.125*** 3.043***  83   (0.03) (0.04) (0.04) (0.04) (0.04)  Number of Individuals 33678 33678 33678 33678 33678  Number of Countries 26 26 26 26 26  Adjusted R2 0.137 0.172 0.173 0.173 0.178  Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Notes: The “contrasts” column indicates significant differences between the survey years, classes, education levels, and party affiliation in models 1 and 2 (*p < .05). 84  Table 3.5 Average marginal effects for class relationships to attitudes towards inequality across years (Model 3) Class 1992 1999 2009   a Working Class  4.164*** b, c  4.186***c  4.149***c  b Intermediate Class 4.064*** a, c  4.165*** c  4.107*** c  c Salariat Class 3.891*** a, b  4.060*** a, b  4.040*** a, b  *** p < 0.001 Note: Letters next to average marginal effects indicate the predicted margins between classes that are significantly  different across years (*p < .05).     Table 3.6 Average marginal effects for party affiliation’s relationship to attitudes towards inequality across years (Model 4) Party 1992 1999 2009   a Left  4.225***  b, c 4.265*** b, c 4.251*** b, c b Centre 4.064*** a, c 4.122*** a, c 4.084*** a, c c Right 3.794*** a, b 3.981***  a, b 3.917***  a, b *** p < 0.001 Note: Letters next to average marginal effects indicate the predicted margins for party affiliation that are significantly  different across years (*p < .05).   Table 3.7 Average marginal effects for class and party affiliation across years (Model 5)  Political Class Groups 1992 1999 2009 (a) Working Class Left 4.278*** c 4.282*** 4.236*** Working Class Centre 4.139*** c 4.120*** b, c 4.136*** c Working Class Right 4.059*** b, c 4.115*** b, c 4.083*** b, c (b) Intermediate Class Left 4.256*** c 4.261*** 4.252*** Intermediate Class Centre 4.104*** c 4.228*** a, c 4.076*** Intermediate Class Right 3.801*** a, c 3.984*** a, c 3.943*** a, c (c) Salariat Class Left 4.153*** a, b 4.269*** 4.272*** Salariat Class Centre 3.951*** a, b 4.035*** a, b 4.041*** a Salariat Class Right 3.532*** a, b 3.831*** a, b 3.750*** a, b *** p < 0.001 Note: Letters next to the average marginal effects indicates the predictive margins between the political classes that  are significantly different across years (*p < .05). 85   Figure 3.3 Linear predictions of inequality views by class (1992, 1999, 2009) 86  Figure 3.4 Linear predictions of inequality views by party affiliation (1992, 1999, 2009)  87  Figure 3.5 Linear predictions of inequality views by class and party affiliation (1992, 1999, 2009)  88  3.10 Discussion The recent economic downturn has generated both academic and public interest in the relationship between economic conditions and attitudes towards inequality. Much of the scholarly work, however, has focused on general patterns in public opinions towards inequality (Brooks and Manza 2013; Kenworthy and Owens 2011; Kenworthy and Owens 2012). Less work has focused on whether stagnating economic conditions provoke critical views towards inequality and whether these are disparate across class and political groups. Focusing on aggregate trends can mask group-based differences in responses to economic change. Social classes, for example, may respond differently to economic shocks while country levels factors remain stable.  Yet, the recent recession reveals the relevance of social class as evidenced by the popular rhetoric of political division between the “99%” and the “1%”. To explore whether economic change incites class antagonisms, I leveraged temporal variation in attitudes towards inequality to examine whether TGR influences class differences in these attitudes and what role political ideology plays in this relationship. How did TGR affect individuals’ attitudes towards inequality? First, the analyses indicate that responses to TGR are patterned by social class. Second, evidence of class-specific responses to TGR suggest that class-convergence may be increasing. Finally, following TGR, class differences in viewing inequality critically varied less by party affiliation.   3.10.1. Class Responses to Economic Crises  The results reveal evidence of the impact that social positioning has on individual attitudes towards inequality during a period of economic instability. A common view in public opinion research suggests that individuals are largely unresponsive to changing economic conditions (Bartels 2013; Brooks and Manza 2013; Cavaillé and Trump 2015; Jetten et al. 2017; Laurin et al. 2013; Trump 2017). Indeed, individual attitudes are often described as “broken” (Brooks and Manza 2013), “weak” (Kenworthy and Owens 2011) or “transient” (Margalit 2013: 83), which suggests that it is likely that individuals are fragmented in their responses to economic recessions. While this may be true in the aggregate, it is more 89  complicated when social positioning is taken into consideration. The analyses revealed patterns of homogeneity in how individuals respond to economic recessions. For example, I found that there are similarities in views towards inequality according to class. TGR did not raise critical awareness of inequality across classes.  A possible reason why this is occurring for the working class is that they may have had it so bad for so long that they begin to feel numb to the effects of economic precarity and instability, even after a shock to the economy. While the vulnerability of the working class in general is a long-term trend predating the crisis, it is important to separate out the implications that general trends and particular shocks have on preferences towards inequality. The economic vulnerability of members of more disadvantaged classes predates the crisis and so this event may have less of an impact on how they view inequality. The limited impact of TGR on working class condemnation towards inequality may also be because the time-period (one year after the recession) could have been too short to catch the TGR effects. Certain segments of the labor markets are exposed to longer terms effects of shocks linked to declining trade barriers, structural changes in the labor force (in some cases imposed by governmental authorities), and the exporting of jobs to low-wage regimes.  The elites, on the other hand, may be sheltered from the negative effects of economic instability because of their secure privileged class position (Modrek and Cullen 2013). The upper classes are less exposed to threats to their jobs because of the compensation agreements characteristic of classes encompassed under a “service relationship”, which offer greater job security and protections during periods of economic uncertainty. With increased protection against economic uncertainty they may have less reason to shift their views towards being more critical of income inequality after TGR and instead preserve the status quo. What these findings suggest is that the general pattern of increasing economic instability may matter more than a specific shock.  Much of the research on public opinions towards inequality points to class-based polarization (Andersen and Curtis 2012; Dodson 2017). The results from this study, however, point towards class 90  convergence. That is, inequality views may be coming together such that the working class differs less from the salariat class. The findings show increasing critical inequality views across classes from 1992 to 1999 and then a decrease from 1999 to 2009. This finding may have policy implications. For instance, this union of views may mean that political backing of social policy geared towards reducing inequality may increasingly include members of more advantaged classes; rather than through grass-roots political movement among the disenfranchised.  3.10.2. Economic Shocks and Class Politics The processes through which economic recessions induce inequality views is one which includes both the differences in terms of the impact of economic conditions on classes and ideological predispositions associated with party differences on economic policy. These perceptual differences, however, arise from the same distinctions which explain differing levels of severity of the recession’s impact: the location of the individual in the class structure. The results from this study reveal how the interplay between class and politics on perceptions of inequality is affected by an economic shock. My findings indicate that party affiliation matters little in determining views towards inequality of the working class in times of crisis. Overall, the estimates indicate that at the height of TGR, party affiliation, was not a driving force behind working class concerns over income inequality. For the working-class, then, shocks cuts across party lines. In comparison, party affiliation plays a greater role in contouring inequality views among the elites. It is in the more advantaged classes, that we see the political differences emerge, i.e. they are more pronounced in the middle and greater still in the upper classes. It is often suggested that as the elites and working classes drift further apart in terms of material inequality, politics will become more divisive (Mayer 2014). Certainly, working-class voters may have moved politically to the right after the crisis in many regards as seen in the surge of various forms of populism (Trump and Brexit); rather than supporting progressive politics (Iversen and Soskice 2015; Luttig 2013; Milojev et al. 2015). During these periods, it is the working classes who share bipartisan views towards inequality in terms of returning to the traditional right-wing stance which is more 91  accepting of inequality. In fact, the only class to see their critical inequality views reach their peak after TGR were salariats who are politically centrist or left leaning.  What might explain this occurrence? Because the impacts of economic recessions are distributed more widely across the class structure (although to differing degrees of severity), individuals who have been previously sheltered from the effects of economic instability—typically those in more advantaged class positions—may shift their attention to economic issues. During periods of economic instability, individuals in more privileged class positions may remain largely sheltered from its negative effects but may become more aware about the possible implications of economic instability from the mass media, which often overreports negative economic conditions. Taken together, these findings point to a different theoretical and analytical approach going forward in the literature on perceptions of inequality. The class-based responses to economic recessions raises issues for recent research on public opinion, in which a central focus has been political polarization over the past few decades (Evans and Pickup 2010; Iversen and Soskice 2015; Whiteley 2016). Scholars contend that public opinion responses to recessions are largely connected to politics. For example, research suggests that individuals move in the conservative direction in response to economic threats (Payne 2017). Studies in this area, however, focus on trends in political affiliation but do not consider how this is affected by social class. Rather than focusing on single sources of influence (economic interest or political ideology) or maintaining that one influences the other, the two aspects must be thought as having varying impacts on perceptions of inequality over time, depending on the changing societal circumstances. Future research would do well to explore the presence of multiple and at times conflicting influences on people’s attitudes towards inequality and focus on identifying the conditions and timing under which form of influence gains prominence over the others.    92  Chapter 4: Class and Country: The Interplay Between Social Class and Societal-level Factors in Determining Citizens’ Attitudes Towards Inequality 4.1. Introduction Rising inequality is a well-documented empirical reality (Alvaredo et al. 2013; Atkinson et al. 2011; Piketty 2000; Piketty 2014). While most people do appear to take notice of inequality and are often critical of it, explanations of the degree of criticism towards inequality remain mixed. Existing research has emphasized macro level factors such as the actual-level of inequality, economic growth and regime type. Research indicates, for instance, that differences in attitudes towards income inequality are partially explained by the actual level of income inequality in a society (Andersen and Yaish 2012). In particular, scholars suggest that awareness of inequality is strongest when inequality is high (Sealey and Andersen 2015). Additionally, scholars find that attitudes towards inequality are the result of countries being at different stages of the modernisation process (Curtis and Andersen 2015). Inglehart’s post-materialist thesis maintains that at high levels of economic development critical appraisals of income inequality should diminish (Inglehart and Flanagan 1987).  Consistent with this argument is the idea that acceptance of income differences should increase as countries accept free markets and democracy (Inglehart and Welzel 2005). Scholars point out that it is important to acknowledge that different political contexts could result in differing levels of public awareness of inequality (Kulin and Svallfors 2013; Saar 2008; Sachweh and Olafsdottir 2012; Svallfors 1997; Taylor-Gooby, Hvinden, Mau et al. 2019). In this vein, many comparative analyses on attitudes towards inequality are based on Esping-Andersen’s (1990) welfare regime approach, which assumes that cross-national differences in support for inequality are impacted by the different institutional frames of welfare regimes. Relatedly, the impact of communism in Eastern Europe is thought to have left people with strong egalitarian preferences despite the fact that most have adopted Western economic practices (Ignácz 2018). Indeed, citizens of communist countries have been found to be much more egalitarian than those of capitalist countries (Kelley and Evans 1993).   93  However, this growing body of literature exploring perceptions of inequality has limitations. Attitudes towards inequality are also affected by an individual’s social position, yet surprisingly few empirical studies have explored the interplay between individual characteristics and broader factors in determining preferences towards inequality. Depending on context, personal experience can increase or decrease discontent with current levels of inequality. For example, Inglehart predicts that periods of economic prosperity will not only reduce criticism of inequality generally, but also coincide with the decline of social class in shaping value preferences (Inglehart and Abramson 1994; Inglehart and Abramson 1999). However, research looking at public opinions towards inequality at the macro level and research into the behaviour of individuals has for many years existed in separate worlds (Anderson and Singer 2008). The growing material differences between classes may lead to similar gaps in the classed perceptions of these differences (Wodtke 2016). Individuals who reside in different positions within the social relations of production might be expected to behave and think in differing ways when it comes to income inequality. The connection between social class and national context could shape perceptions towards inequality. For instance, critical attitudes towards inequality among classes may differ in contexts of high or low inequality because the impact of the actual reality of inequality in these contexts varies according to class. Paper three begins to parse out this possible relationship by exploring the degree to which social class interacts with societal-level aspects to influence attitudes towards inequality.  Do perceptions of inequality of the dominant and dominated fractions of society respond similarly to shifts in objective inequality? Do other factors at the country-level play a role in shaping peoples’ perceptions of inequality alongside social class? This chapter has at its focus how broader influences at the societal level variously impact individual perceptions of inequality for people in different class positions. In this chapter, I integrate individual attributes and broader macro-level differences across countries to analyze citizens’ value preferences towards inequality. More specifically, I focus in on the 94  interplay between how countries’ macro characteristics and differences among the economic position of individuals together shape public opinion towards inequality over time and place.  4.2. Class and Country The logic underlying this chapter is that individuals form attitudes in variable environments and these environments can come in the form of differential economic, social and political conditions that may shape individual interpretations and actions. People receive different types of information from the countries where they live, and this contextually supplied information shapes a plethora of political attitudes and behaviours. In this way, citizens do not form opinions towards inequality in a vacuum. At the societal level scholars suggest that objective inequality (Andersen and Curtis 2012; Curtis 2016; Kenworthy and McCall 2008), economic growth (Inglehart and Abramson 1994), regime type (Sachweh and Olafsdottir 2012), as well as the shift from a communist to a capitalist state (Inglehart and Welzel 2005; Kelley and Zagorski 2004; Koçer and van de Werfhorst 2012) are important determinants in how people perceive inequality. In the following section, I detail how the relationship between social class and citizens’ views towards inequality may be influenced by these national-level factors.  4.2.1. Economic Dispersion and Views Towards Inequality Are citizens’ collective perceptions of inequality reflective of the objective reality of the income gap? The extent of material inequality in a particular society may shape public opinions towards the income distribution. For instance, citizens in countries with higher levels of income inequality might be expected to express more negative attitudes towards inequality (Anderson and Singer 2008). This makes sense because if the level of income inequality is high, people are exposed to more of its negative effects, and this may increasingly register in their consciousness.  Attitudes towards inequality seem to be affected by the actual levels of inequality in the societies in which people live (Curtis and Andersen 2015; Curtis 2016; Niedzwiedz and Kandlik-Eltanani 2014). Indeed, research shows that support for redistribution and other social policies connected to equality 95  increases with actual levels of inequality (Kulin and Svallfors 2013). Increasing inequality may encourage people to be more critical of inequality and become aware of various issues related to wealth and redistribution. On the basis of previous research, I test the following hypothesis: Hypothesis 1: National income inequality, as measured by the Gini coefficient, is positively associated with inequality views. That is, people living in less equal societies are more likely to be critical of inequalities in their society than those living in more equal societies.  Conversely, other scholars suggest that there has been little public reaction to these trends of increasing inequality and their consequences (Osberg and Smeeding 2006). The reality of inequality and negative opinions towards it, often do not match up. Research in the United States suggests that inequality has grown without any noticeable public resentment, which represents a challenge to the notion that increasing objective inequality is matched by a corresponding increase in critical views (Andersen and Fetner 2008; Brooks and Manza 2013). For example, McCall (2016) contends that in the United States concern about inequality corresponds more to the perceptions of the consequences of inequality rather than with the actual level of inequality. Kenworthy and McCall (2008) also demonstrate this variability by showing that objective inequality does have an influence on perceptions of inequality in some countries (e.g. Sweden, Italy, Australia, United Kingdom) but not others (Norway, Germany, Canada). Evidence of whether a connection exists between the actual level of inequality and citizens perceptions is thus mixed and far from conclusive. A lack of clear relationship between subjective and objective inequality suggests that additional factors may shape the degree of resentment towards inequality.  As a small prelude, Figure 4.1 illustrates the link between the actual level of inequality and public acceptance of inequality. Income inequality is plotted on the x-axis and the country mean score of perceptions of inequality on the y-axis. Higher scores indicate greater disapproval of income differences. The figure reveals that few citizens are able to accept income differences. It also shows no discernable relationship between income inequality and citizens’ critical views towards it. On the surface, this appears 96  to support national studies which show no clear link between the level of inequality and discontent (Larsen 2016; McCall and Kenworthy 2009).  Source: Authors calculations using International Social Survey Programme data 1992, 1999, 2009.   Figure 4.1 Country averages on subjective inequality index by income inequality (Gini) 97  4.2.2. Economic Development and Views Towards Inequality For others, the cause of differentiation in attitudes towards inequality is economic development (Inglehart and Rabier 1986; Inglehart and Flanagan 1987; Inglehart and Welzel 2005). Inglehart’s (2005) theory implies that the absence of a connection between viewing inequality critically and the level of inequality is a matter of countries being at different stages of the modernization process. The basic idea here is that acceptance of inequality will increase as free markets and democracy generate more and more prosperity. A country’s economic affluence should trigger less public condemnation of inequality. This theory implies that economic position has little impact on intolerance towards inequality because even the most disadvantaged are comparatively free from material concerns.  Motivated by Inglehart’s (2005) claim that economic development contributes to the shift towards post-materialist values in modern societies, I investigate the relationship between economic development (GDP) and perceptions towards inequality. Consistent with Inglehart’s post-materialist thesis, I suspect that there is a negative relationship between economic development and critical perceptions of income inequality. I test this thesis with the following hypothesis: Hypothesis 2: People are more critical of inequality in countries that have relatively low economic development (GDP).  Figure 4.2 charts the level of resentment towards inequality against GDP for various countries. With rising prosperity, disapproval towards inequality decreases. The figure shows a negative relationship between economic growth and viewing inequality critically. The graph appears to add credence to the claim made by Inglehart that more favourable views towards inequality coincides with economic growth.  At the same time, both Figures 4.1 and 4.2 reveal that there is considerable variation between countries in terms of the predictive capacity of income inequality and economic development on citizens’ negative views towards the income distribution. What might account for these cross-national differences 98  in disapproval of inequality? Is the relationship between objective inequality and economic growth on citizens’ subjective views towards inequality uniform?  Figure 4.2 Country averages on subjective inequality index by economic growth (GDP) Source: Authors calculations using International Social Survey Programme data 1992, 1999, 2009.   99  4.2.3. Theorizing the Interplay between Class, Objective Inequality and Economic Growth on Attitudes Towards Inequality It seems reasonable to suggest that the effects of economic contexts on attitudes towards inequality are either strengthened or weakened depending on an individual’s placement within the social relations of production. The fundamental argument is that the economic situation of the individual can either shield them from inequality or provide a glimpse of it. For instance, some learn about inequality during the course of their education, while others learn about it through their exposure in various life settings, workplaces for instance. Those occupying disadvantaged class positions may be more aware of existing inequalities in their society. Disadvantaged classes are generally more likely to see unequal distributions in their society as unjust regardless of the reality of inequality. The most disadvantaged segments of the population might be expected to perceive inequality critically irrespective of the actual level of inequality. Because of their prolonged exposure to the negative effects of inequality, working class displeasure with inequality may endure regardless of shifting patterns of inequality at the national level. For those who have not been exposed to the negative effects of inequality (perhaps those residing in positions of privilege) macro-aspects, such as the actual level of inequality, may step into inform their views towards inequality. Increased awareness of the objective changes in the real economy may heavily condition the views the advantaged classes have towards inequality. The relationship between social class, inequality and critical views towards the income distribution suggests that inequality and class-related aspects interact to influence perceptions of inequality (Andersen and Fetner 2008; Carroll, Casswell, Huakau et al. 2011; McCall and Percheski 2010; McCall and Manza 2011). The logic behind this is at those who are in more precarious class positions tend to have much more to gain from increasing redistribution; whereas, those in the most secure and prosperous class positions tend to benefit from large income differences and thus are less supportive of equality. Building on this research as well as hypothesis 1, I investigate how class interacts 100  with national-level inequality to influence perceptions towards income inequality. I pose the following hypothesis: Hypothesis 3: Differences in attitudes towards inequality among the class fractions of society will be greatest when inequality is high. In opposition, Curtis and Andersen (2015) find that when income inequality is low, those in lower economic positions tend to be more critical of inequality in comparison to those in higher economic positions who favour increases in inequality. However, when income inequality is high the “middle” and “upper” classes are just as likely as the working classes to be critical of inequalities. That is, when inequality rises more people across the class structure are more critical of inequalities, and attitudes for the different social classes may converge (Curtis and Andersen 2015). In this way, critical views towards inequality may be more polarised along class lines in countries with low income inequality; yet these may come together in class terms when inequality rises. The idea here is that as inequality rises, class divisions become clearer and so individuals, regardless of their class, are better able to see the extent of material inequalities produced by such divisions, which leads to a “rising tide” of critical perceptions towards inequality across the class spectrum. Rather than persistent class divisions in perceptions of inequality, increasing levels of objective inequality may encourage all classes to be more critical and aware of various issues related to wealth and redistribution. To test this competing theory to hypothesis 3, I offer the following hypothesis: Hypothesis 4: The gap between those in more advantaged class positions and those in the working-class occupations in viewing inequality critically will be wider when country-level income inequality is lower, but this gap will significantly attenuate when income inequality is higher. Scholars also argue that economic development and modernisation have reduced the importance of social class with regards to people’s perceptions (Clark and Lipset 1991; Inglehart and Welzel 2005; Shibutani 1955). The post-materialist thesis generates two expectations with respect to perceptions of 101  inequality: 1) countries with highly developed economies should exhibit relatively lower levels of resentment towards inequality compared to developing countries; and 2) because citizens of highly developed countries enjoy considerable wealth and affluence, post-materialism should be diffuse, which suggests that class should have little impact on intolerance towards inequality in countries where economic growth is higher.  There are, however, reasons to believe that post-materialism is less widespread than its advocates suggest. Those who claim that modernisation and economic development induce favourable sentiments towards society often do not recognize that national-level economic prosperity does not operate the same for all members in a society. The theory largely fails to consider that resources are often distributed unequally within society and this prevents it from explaining much of the connection between class and views towards inequality. It could be that class differences in perceptions of inequality remain unchanged as the level of economic development increases (Andersen and Fetner 2008). On the basis of this logic, I test for an interaction between social class and economic development (GDP). To reveal GDP’s potential significance for the relationship between social class and critical attitudes towards inequality, I test the following hypothesis:  Hypothesis 5: The gap between those in more advantaged class positions and those in the working-class occupations in viewing inequality critically will remain unchanged as economic development (GDP) increases. 4.3. The Class Politics of Welfare Regimes  In the face of difficulties in explaining variation in preferences towards inequality using objective indicators such as the extent of material inequality, other institutional approaches have gained footing. Arguments exist that varying production regimes shape individual preferences towards income inequality. A dominant theme has emerged in the literature that the particular welfare regimes described by Esping-Andersen (1990) influence acceptance of inequality. The regime typology developed by Esping-Andersen 102  outlines the processes in which different states’ social policies have formed into three distinctive types: liberal, social democratic, and conservative6. The different kinds of institutional structures associated with the various regimes may be operative in transmitting certain social values to citizens and these may be important in relation to variation in public opinions towards income differences.  Social democratic regimes tend to have social policies geared towards attempting to shelter individuals from the deleterious effects of the markets and inequality. The relative generosity of welfare states may generate a “feedback mechanism” whereby, the social policies which protect individuals from poverty may generate public support against income inequality (Larsen 2008: 155). In comparison, liberal regimes tend to follow a neoliberal trajectory in which states intervene minimally in the market and stress individual responsibility. The repercussions of neoliberal rationality on social policy may lead to public opinions which are more favourable of income inequality. Indeed, scholarship points towards the persistence of these institutional cultures over time. Scholars find that these have lingering impacts on public opinions even after a change in regime type (Jepperson 1991). Further, social democratic regimes originated out of left-wing labour parties and so have stakes in universal social rights and more equitable social policy ideals. Both liberal and conservative regimes trend to favour right-wing or right of centre politics, which are geared towards meritocracy. For conservative regimes, state responsibility for individual welfare is greater than in liberal regimes; while equity is stressed it is predominantly done so for security reasons.  For Mediterranean regimes, preferences for redistribution are thought to be similar to conservative regime types (Arts and Gelissen 2002; Gelissen 2000). What this elucidates is that citizens’ opinions on a vast area of social policy preferences may be reflective of the type of welfare state.  The influence of this regime typology in shaping public opinions is well-supported. For instance, scholars find that there are significant variations in redistributive policy preferences based upon                                                      6 More recently, the three-pronged typology was broadened to include Mediterranean and transition countries (see Arts 2002) 103  institutional contexts (Esping-Andersen 1990; Larsen 2008; Svallfors 1997). Svallfors (1997) has shown, for instance, that as measured by attitudes towards wage differences, citizens of social democratic welfare regimes do appear to be more egalitarian; while there is no clear distinction found between liberal or conservative welfare regimes. To determine whether there is a relationship between regime type and inequality views I test the following hypothesis: Hypothesis 6: Individuals in countries characterized by a social democratic regime type should be more critical of inequality, in comparison to those living in liberal and conservative regime types.  The various regimes also differ in terms of the prevalence of class discourse. One theory is that the classed-interpretations of society in social democratic regimes may impact the connection between class and critical views towards inequality (Wright 1997). Wright (1997) asserts that the political discourse in social democratic welfare regimes often imbues class politics. Social democratic regimes, are regarded by most scholars as “having historically prominent and well-established class cleavages” (Brooks and Svallfors 2010: 200). Indeed, Svallfors (2002) suggests that the welfare state represents one of the most important arenas for class politics. In this way, the classed-interpretations of the social world found in social democratic welfare-state regimes, may strengthen the relationship between class and inequality views. Another theory posits that welfare-state policies may, instead, mitigate the class-risks that people are exposed to; making the link between perceptions of inequality and class positions less salient (Kulin and Svallfors 2013; Svallfors 1997). The existence of redistributive institutions may modify the way members of particular classes make connections between interests, values, and inequality (Kulin and Svallfors 2013). Generous welfare-states may buffer disadvantaged classes against social risks, generating a sense of economic security (Anderson and Hecht 2014). To test these competing theories, I pose the following hypothesis: 104  Hypothesis 7: Patterns of class polarization in viewing inequality critically are stronger in societies characterised as social democratic in comparison to liberal or conservative regimes.  4.3.1. The Spectre of Communism and Views Towards Inequality  A further institutional distinction must be made between market-oriented societies and former communist societies. This is because in former state socialist societies a common finding is that people hold more egalitarian attitudes regarding income inequality (Gijsberts 2002; Kelley and Evans 1993). Indeed, scholars have shown that citizens of post-Communist nations maintain strong allegiances to social welfare and equality despite holding views more favourable to free markets than in the 1990s (Arts and Gijsberts 1998; Roller 1994). For instance, Gijsberts (2002) finds that people in Hungary and Poland both before 1987 and after 1992, held more egalitarian views towards income inequality than those in market societies. While faith in the communist ideology may be lost in former-communist nations, there remains a deeply rooted preference for government intervention and equality (Breznau 2010). In societies that experienced a shift from total government regulation to market regulation, publics’ internalization of the norms and values of communism may lag behind the actual economic changes occurring. The findings that people in former state-socialist societies often hold more egalitarian views towards income inequality in comparisons to those in market societies develops from the theory of the “spectre” of communism in Eastern Europe shaping individual attitudes (Evans and Kelley 2017; Gijsberts 2002). In other words, individuals in former communist countries have been found to be much more egalitarian and thus may view inequality more critically compared to those of liberal regimes. Following previous research regarding the role of communism on perceptions of inequality, I test the following hypothesis:  Hypothesis 8: The experience of communism is positively related to egalitarian attitudes and thus individuals in former communist societies should be more critical of inequality in comparison to those in liberal regimes.  105  In former communist societies, widespread destratification policies were created with explicit class politics which involved discrimination against privileged social groups in favour of the disadvantaged working classes. The role of communism was not solely to persuade the citizenry that the working class deserves more, but that the elites deserve less. This communist political ideology favoured greater equality for the masses while engaging with explicit class discourse. At the aggregate level, former communist countries may indeed be more critical of income inequality in comparison to market societies; however, at the individual level, differences within former-communist countries may vary more by class due to the historic prevalence of class discourse dispersed throughout former communist countries. In contrast to former-communist countries, the political ideology in liberal regimes has historically undermined class vernacular. For example, in the United States, the Democratic Party has often displaced their political discourse from the language of class. While there have been more recent exceptions (e.g. Bernie Sanders, Jeremy Corbyn), political strategies and ideologies are often organized in non-class ways in liberal regimes (Wright 1997). On this basis, I test the following hypothesis: Hypothesis 9: The class differences in viewing inequality critically are greater in former-communist societies in comparison to liberal regimes. 4.4. Data What the aforementioned literature points towards is that variations in perceptions of inequalities may be connected to both societal-level and individual-level determinants. As such, aspects at both levels must be taken into consideration when exploring what motivates people to view inequality critically across time and place. The analysis to follow accommodates for individual as well as societal-level predictors and models country-level variance to uncover patterns in discontent towards inequality. The study extends research by merging individual-level data on attitudes towards income inequality with country-level data on material inequality, economic growth, and regime type, to analyze whether class differences in attitudes towards income inequality are susceptible to factors at the societal level. A benefit of this cross-national design is in its variation in these contextual factors at both the national and temporal 106  levels. This variation is beneficial in terms of examining how increasing material inequality affects class politics, while also “controlling” for additional confounding sources of political attitudes.   To test my hypotheses, I use individual level data from the ISSP social inequality modules from 1992, 1999 and 2009, as well as national-level indicators (Gini, GDP and regime type) obtained from the World Income Inequality Database, the Standardized World Income Inequality Database (SWIID), and the United Nations Statistical Division Aggerate National Accounts. The ISSP consists of nationally representative sample of adults aged 16 years and older administered over four waves in 27 countries over a 17-year period. To account for the diversity of national contexts, I include all countries for which data are available.7 My analysis thus utilizes data collected on 24 countries (8 of which were measured in all three iterations) in 1992, 1999 and 2009. Details on the descriptive information for each country and year are listed in Table 4.1.                                                         7 Chile and the Philippines are omitted as their values on the Gini were outliers. Israel is omitted as it is considered “unclassified” under the regime typology.  107  Table 4.1 Descriptive statistics for each country and year Country Survey Year GDP Per Capita* Gini Regime Type Mean Subjective Inequality Index Australia 1992 7040 .331 Liberal 3.62  1999 9366 .381  3.76  2009 12638 .348  3.91 Austria 1992 2751 .275 Conservative 4.09  1999 3265 .26  4.23  2009 3845 .262  4.3 Bulgaria 1992 342 .338 Former Communist 4.78 1999 308 .326 4.78  2009 499 .289  4.5 Canada 1992 10024 .283 Liberal 3.81  1999 12766 .359  3.76 Cyprus 1999 179 .299 Mediterranean 3.65  2009 252 .321  3.73 Czech Republic 1992 1267 .181 Former Communist 3.9  1999 1452 .257  4.4  2009 2024 .181  4.28 France 1999 22589 .29 Conservative 4.4  2009 25958 .29  4.56 Germany 1992 27518 .291 Conservative 4.23  1999 30340 .299  4.04  2009 32831 .20  4.37 Hungary 1992 890 .282 Former Communist 4.2  1999 1028 .292  4.58  2009 1300 .247  4.74 Italy 1992 17909 .334 Conservative 4.38  2009 20898 .334  4.59 Japan 1999 52043 .303 Liberal 3.9  2009 54707 .311  4.11 Latvia 1999 156 .333 Former Communist 4.52  2009 247 .333  4.52 New Zealand 1992 824 .316 Liberal 3.89  1999 1093 .33  3.88  2009 1452 .375  3.69 Norway 1992 2279 .241 Social Democratic 3.75  1999 3556 .273  3.79  2009 4262 .24  3.52 Poland 1992 2161 .324 Former Communist 4.18  1999 3120 .291  4.32  2009 4626 .314  4.36 108  Portugal 1999 2133 .36 Mediterranean 4.75  2009 2339 .341  4.54 Russia 1992 11499 .397 Former Communist 4.37  1999 8647 .374  4.71  2009 14592 .357  4.54 Slovak Republic 1992 407 .245 Former Communist 4.29  1999 548 .197  4.65  2009 852 .245  4.52 Slovenia 1992 266 .227 Former Communist 4.26  1999 355 .225  4.36  2009 474 .227  4.51 Spain 1999 10917 .322 Mediterranean 4.22  2009 14314 .312  4.19 Sweden 1992 3137 .23 Social Democratic 3.6  1999 3786 .261  3.87  2009 4608 .23  3.93 Switzerland 2009 5668 .317 Liberal 4.2 UK 1992 16307 .324 Liberal 4.08  1999 20211 .32  4.05  2009 24005 .339  3.97 US 1992 93698 .379 Liberal 3.92  1999 122133 .354  3.76 2009 145948 .381 3.76 *GDP is divided by 100,000,000         109  4.4.1. Dependent Variable  The dependent variable has been recoded from the following survey question: “How much do you agree or disagree with the statement ‘Differences in income in [respondent’s country] are too large?’”. Response categories include: strongly agree (1), agree (2), neither agree nor disagree (3), disagree (4), and strongly disagree (5). Following previous research, I recoded the variable into an inequality index with the following values: ‘‘agreed strongly’’ 5, ‘‘agree’’ 4, ‘‘neither agree nor disagree’’ 3, ‘‘disagree’’ 2 and ‘‘disagree strongly’’ 1.  4.4.2. Individual-level Variables  I employ two measures of individual social position: social class and education. As with the previous chapters, class is operationalized using the European Socioeconomic Classification (ESeC) (Rose and Harrison 2007; Rose and Harrison 2010). I use the 3-class version which consists of the working, intermediate and salariat classes, with the latter as the referent category. Status is measured though education which is coded into the following categories “Less than secondary”, “Secondary/higher secondary”, “Above higher secondary” and “University degree” as the referent category. Party affiliation is measured categorically with “Left”, “Centre”, and “Right” as the reference group.8 All models also control for gender, age, marital status and religion.  4.4.3. National-level Variables 4.4.3.1. Income Inequality  The data on perceptions of inequality are matched with the data on the actual extent of income inequality in a particular society. There are various measures of income inequality; however, the most                                                      8 Analyses were also carried out with non-voters included. The results can be found in table C.3 in the Appendix. The technique (Stata command “Suest”) used to test the differences across models in the previous chapters is not supported for mixed effects models. Nevertheless, by visually comparing effects size and levels of significance we can see that the results between the two models are fairly similar for key variables besides the following: in the “non-voter” model the intermediate class coefficient in model 4 is significant and the working class by conservative interaction turns insignificant. As with the previous empirical chapters, the results here pertain only to those politically engaged. 110  common is the Gini coefficient. The Gini measures income inequality on a scale from 0-1. A perfectly equal society would be one where the Gini coefficient is 0. Extreme inequality would be represented by a society where the level of inequality equates to 1. The main sources for the Gini are the World Income Inequality Database (https://www.wider.unu.edu/project/wiid-world-income-inequality-database). The Gini is measured using disposable income, post taxes and transfers. This measure of inequality takes into account the dispersion of income while considering the effects of taxes and social spending in a particular country. The difference in Gini indices across the countries using the post taxes and transfers measure is generally narrower than when the Gini is measured using market incomes. Missing country values were obtained from the Standard World Income Inequality Database (Solt 2009). The Gini was centred before entering it into the statistical models. The Gini may not be the best measure of objective inequality. For one, it does not account for the wealthiest one percent of the population, who are often excluded from household surveys. Using the Gini coefficient as a measure of income inequality can be problematic, with national accounts often providing a greater understanding of the incomes of the elite (Milanovic 2013; Székely and Hilgert 2007). Indeed, Chandy (2018) has shown that when this is accounted for, the Gini rises by 9 percentage points. Their new calculations indicate that inequality is, on average, much higher than is reported. Nevertheless, the Gini represents the best available measure of objective inequality when conducting comparative research using a number of countries.  4.4.3.2. Economic Development Economic development is measured by gross domestic product (GDP) per capita, standardized to 2010 U.S. dollars for each survey year. GDP per capita in US dollars was obtained from Aggregate National Accounts (https://unstats.un.org/unsd/snaama/dnlList.asp). GDP was logged before entering it into the models as the distribution of the raw GDP variable was skewed to the right. In order to make the results of the models more easily interpretable, I divide GDP per capita by 100,000,000.  111  4.4.3.3. Regime Type  To measure regime type, I code the countries according to Esping-Andersen’s (1990) regime typology. The clustering of welfare regimes corresponds to the conventions of research previously undertaken in comparative welfare regime analyses (Dallinger 2010; Gelissen 2000). Accordingly, Sweden and Norway belong to the social democratic regime type. France, Austria, Italy and Germany are considered conservative. Great Britain, the United States, Canada, Australia and New Zealand are grouped as liberal regimes. Spain, Cyprus and Portugal count as Mediterranean types. The former communist countries are Bulgaria, the Czech Republic, Hungary, Latvia, Poland, Russia, Slovak Republic, Slovakia and Slovenia. Japan is ambiguous but is usually categorized as a liberal regime (Dallinger 2010). In the statistical models, regime type is entered as dummies into the estimation, with “liberal” as the referent category.  4.5. Analytical Strategy  What is the importance of aspects at the individual and societal levels for critical views towards inequality? Since I am interested in attitudes towards inequality within countries as well as between them, I employ a set of multilevel models to test the hypotheses. This multilevel approach assumes that individual citizens are nested in larger cross-national macro environments that shape individual behaviour but that this impact is also moderated by citizens’ individual characteristics in terms of economic positioning.  By employing a multilevel approach, I am able to include both individual and country-level predictors of attitudes to inequality. This is necessary as the substantive focus of this chapter is on how country-level factors interact with class in its relationship with inequality views. Having developed the relevant individual-level factors that contribute to perceptions of inequality in Chapter two as well as the trends in these perceptions in Chapter three, I can now develop and bring together these into a model and introduce country-level predictors.  112  The multilevel model describes the relationship between the dependent variable and explanatory variables with coefficients that can vary on one or more of the higher-level units. Perceptions of income inequality are shaped by societal and individual level differences but also by unobserved country differences. The approach here pools the data as in the first two papers, but rather than treating country effects as separate values to be estimated, these are modelled as random draws with a variance which is estimated (Bryan and Jenkins 2016). Under the random effects models, the unmeasured effects that can vary across countries but not across time are modelled with country-specific error terms. 9 By employing a mixed effects model I am able to account for additional sources of confounding at the country-level by estimating the variability in these effects. Including country-year variance, enables me to make inferences about this variance and whether it is significant. Further, as class may operate differently in different countries, I include a random slope for class which assumes that the relationship between class and inequality views differs across contexts. By including a random component for class, I am able to account for its distinct within-country and between country effects, which allows for a more substantive interpretation of the relationship between class an inequality views (Heisig and Schaeffer 2019). All models also include year dummy variables.  I begin by fitting Model 1, which includes the contextual variables alongside the individual-level predictors. 10 This model offers tests of hypotheses 1, 2, 6 and 8 which contend that subjective inequality is affected by the objective reality of income inequality, economic development and regime type. Model 2 includes the interaction between class and Gini in order to adjudicate between hypotheses 3 and 4, which offer competing theories about the relationship between class and inequality views alongside the actual                                                      9 In order to determine whether to include a random effect for the particular country-year covariates within the context of a multilevel modelling approach, I ran a Breusch-Pagan Lagrange multiplier (LM) test. The null hypothesis in the LM test is that variances across country-years is zero. In other words, no significant differences across country-years. The results of the test are significant (p < .001) so I reject the null hypothesis and conclude that a random effects approach is appropriate. In other words, there is evidence that the countries vary with respect to average attitudes to inequality, which means that I should account for this by modeling that variation. This test offers statistical support for including a random effect for the country-year covariates.  10 As a preliminary analysis, I explored the nature of associations between variables across levels 1 and 2. See table C.1 and the related description in the Appendix.  113  extent of inequality in a society. This model allows me to compare the relationship of class to attitudes towards inequality across levels of national inequality while accounting for heterogeneity in both average levels of disapproval of inequality and effects of class at the country level. Model 3 adds an interaction between class and GDP which offers a test for hypothesis 5, which contends that the relationship between class and inequality views remains unchanged irrespective of economic development. Model 4 includes an interaction between class and regime type, which tests hypotheses 7, 8 and 9 that class operates differently depending on the regime typology of the country. All models are approximated using full maximum likelihood and employ probability weights. 11 . Further, all models employ an “identity” covariance structure which assume equal variance for the random effects and all covariances are fixed to 0. Are there significant differences between classes in particular regimes, or are the differences between regimes in their levels of actual inequality explaining the different inequality views of the classes? Countries characterized as social democratic regimes are also countries where the actual levels of income inequality tend to be lower than in liberal regimes (OECD 2016). It is possible then that the level of inequality within particular regimes is generating stronger class differences in inequality views rather than the characteristics of the regimes themselves. Gini could be a confounder, meaning that the relationship between class and regime type could be driven by differences in income inequality across regimes. To explore this, in Model 5 I include an interaction term between class and Gini in addition to the class and regime type interaction, to uncover whether it is the objective reality of inequality in various regime types which may confound the relationship between class, regime type and inequality views.   4.6. Results To begin, I look at Model 1 in Table 4.2. Model 1 reveals whether national-level factors play a role in shaping disapproval towards inequality. The results show that neither Gini nor GDP are significantly                                                      11 A noted limitation is that probably weighting is applied at level one only, which means that I assume countries have equal probability of selection.  114  associated with viewing inequality critically. This result refutes hypothesis 1, which assumed that those living in less equal societies are more critical of inequality compared to those in more equal societies. Based on the lack of statistical significance between GDP and critical perceptions of inequality, hypothesis 2 must also be rejected. The objective reality of inequality and economic development have no overall relationship to predicting disapproval towards inequality.  Shifting now to regime type, the results show that certain regimes correspond to more critical views towards inequality on average. Citizens in social democratic regimes tend to be less critical than those living in liberal regimes, but this is not a significant difference. These results go against hypothesis 6, which posited that citizens of social democratic regimes should be less critical of inequality in comparison to those in liberal regimes. Former communist regimes appear to be the most critical of inequality relative to liberal ones in that they view inequality .53 points higher on the index, all else equal. In terms of the social class, we see that the working class are .16 higher on the inequality index relative to the salariat; while the intermediate class are .11 higher. The contrast column next to Model 1 indicates the significant differences within the categories of the key variables. This column reveals that there are significant differences between classes and across the political spectrum. In terms of regime type, we see that social democratic regimes are significantly different from all other types besides liberal ones.  Up to this point in the analysis, we have looked solely at the independent impact that various contextual-level factors and class have on critical views towards inequality. I now shift to look at their interplay. Model 2 includes an interaction between class and Gini to explore whether classed disapproval towards inequality fluctuates with the actual level of inequality in societies. The significant values of the coefficients for class and Gini interactions reveal that the relationship between class and inequality views are partially determined by the objective reality of inequality in a society. We see that the salariat has more critical views towards inequality as the Gini coefficient rises. While the effect of overall country-115  level inequality is in the same direction for the working and intermediate classes, it is significantly less pronounced. Jointly, the interaction between class and Gini are also significant (p < .001).12  Model 2 confirms the significance of income inequality in moderating the relationship between social class and perceptions of inequality. It should be noted that while the slope for the salariat is meaningfully different from the slopes of the working and intermediate classes, the difference between classes may or may not be significantly different for different values of Gini. To illustrate this, in Table 4.3, I provide the class differences in the average marginal effects across certain values of Gini along with the significant between class differences. Significant differences between the classes are found for Gini values less than .35 (or centered Gini values less than .05). In all classes, we see significant differences in attitudes towards inequality at lower levels of inequality that become smaller and no longer significant as inequality rises. This suggests that a cross-level interaction between social class and national context may be important in constructing a theory of perceptions of inequality. To further facilitate the readings of the interactions I have derived the predictive margins of the classes in critically viewing inequality at selected values of the Gini and plotted these in Figure 4.3.  Figure 4.3 reveals new insights on the relationship between social class, national-level inequality, and critical condemnation towards inequality. For one, they disprove hypothesis 3, which emphasized that the differences in critical attitudes among the classes will be greatest when inequality is high. In fact, the results show the opposite pattern: differences in critical views towards inequality among the various classes are smallest when inequality is high. As a result, hypothesis 4 which presumed that the gap between the salariat and the working class in viewing inequality critically should be the widest when income inequality is low is confirmed. The gap between classes is bigger in contexts of lower inequality and there are almost no differences in contexts of higher inequality. As this figure illustrates, class polarization in critically viewing inequality is smaller when the actual level of inequality is high. In                                                      12 Table C.2 in the Appendix provides joint significant tests for Models 1 through 5. 116  circumstances where this is the case, the working classes are actually less critical of inequality in comparison to their counterparts in salariat class positions.  When inequality is higher, all classes declare more condemnation of inequality than when it is lower. However, this effect is stronger for the salariat classes relative to the working classes, such that their views converge. While there is a broader convergence in disapproval towards inequality across all classes as inequality increases, change in the critical views of the salariat classes is the most pronounced. Figure 4.3 indicates that for the salariat classes the effect of increasing country inequality on their critical perceptions increases quite sharply as inequality increases. While all class evaluations become more critical as inequality levels increase, the positive effect of higher levels of inequality is substantially larger for the salariat. This suggests that when inequality is high people across the class spectrum resent inequality; however, this is most marked in the advantaged classes such that in these contexts the salariat classes are more critical than the intermediate and working classes. This may be the case because when inequality is high, broader negative consequences are felt increasingly among the advantaged classes. In lower inequality contexts, by contrast, those at the top of the class system are less likely to be susceptible to the broader negative impacts of inequality.  I now shift gears to look at the interplay between class and economic development.  Model 3 includes an interaction between class and GDP in order to test hypothesis 5, which asserts that as GDP increases, the gap between the salariat and the working class will remain unchanged. Jointly, the interaction between class and GDP is not significant neither are any of the interaction terms between class and GDP. Taken together, these findings reject hypothesis 5. Evidently, economic development does little to reduce the importance of class in its relationship with critical views towards inequality.  In Model 4, I include an interaction between class and regime type. This reveals whether the classed displeasure with inequality is patterned by regime type and whether this is strongest in former communist countries. The interaction term is jointly significant (p < .001). Significant effects are found among the working classes in all but former communist regimes; while significant effects are found 117  among the intermediate classes in social democratic and Mediterranean regimes. In Table 4.4, I include the average marginal effects of the classes across the regime types, including the predicted margins between classes that are significant across regime types. The table reveals that the majority of significant differences between classes are in conservative and social democratic regimes. Between classes in Liberal and Mediterranean regimes, there are fewer significant class differences and within former communist regimes there are none.  To better understand what is going on, in Figure 4.4 I include the predictions from Model 4 in a stacked bar chart alongside the associated confidence intervals. The figure documents that there is greater class polarization in countries characterized as social democratic in comparison to the other types. This is evidenced by the gap in the vertical bars among the various classes. Thus, this supports hypothesis 7, which supposed that patterns of class polarization to be the most evident in social democratic regimes in comparison to liberal or conservative regimes. Figure 4.4 is also revelatory in that it shows that the relationship between class and inequality views seem to be affected by whether a society is formerly communist. Here we see that rancor towards inequality is greatest for all classes in countries characterized as former communist. This also confirms hypothesis 8 which suggested that individuals in former communist societies should be more critical of inequality in comparison to those in liberal regimes. Indeed, inequality views of the classes are similar in former communist societies compared to classes in liberal regimes, which are more different. Taken together, this rejects hypothesis 9, which assumed that the class differences in viewing inequality critically are greater in former-communist societies compared to liberal regimes.  There could be confounding in the interactions between class, inequality and regime type. That is, it is useful to determine whether the regime types are generating the class differences in inequality views, or whether it is the varying levels of inequality within particular regimes which leads to the relationship. For instance, it is possible that lower inequality in social democratic regimes is generating stronger class differences rather than the characteristics of the regime type itself. To unpack whether there is 118  confounding in the interactions between class, Gini and regime type, in Model 5 I included an interaction term between class and Gini in addition to the interaction between class and regime type. Jointly, the interaction term between class and Gini and class and regime type are significant (p <.001). Moreover, jointly the main effects for class remain statistically significant (P < .001). Also important, the Akaike information criterion (AIC) values for Model 5 indicate that this is the best-fitting model.  Table 4.5 provides the class differences in the average marginal effects across Gini with the associated indicators of statistical significance. We see that the general pattern found in model 2 of more significant differences between classes in contexts of lower inequality and fewer in contexts of higher inequality persists. Table 4.6 offers the class differences across regime types when the interaction between class and Gini is included in model 5. The pattern of class differences in inequality views across regime types is similar to that of model 4; although the difference between the salariat and working class in Mediterranean regimes is no longer significant and there are now significant differences between all classes in liberal regimes. This further supports hypothesis 7 that the class discourse associated with social democratic regimes plays a role in generating the class differences in inequality views, but also suggests that this relationship occurs irrespective of the extent of inequality apparent in these societies. 119  Table 4.2 Mixed effects linear models predicting critical views towards inequality (Standard errors in parentheses)  (Model 1) (Model 2) (Model 3) (Model 4) (Model 5) Individual-level      a Working Class (ref: Salariat) 0.155*** b 0.121*** 0.008 0.129** 0.165***  (0.02) (0.02) (0.46) (0.05) (0.05)       b Intermediate Class 0.106*** a 0.083*** 0.094 0.068 0.088*  (0.02) (0.02) (0.40) (0.04) (0.04)       c Less than secondary 0.210*** e 0.210*** 0.210*** 0.212*** 0.212*** (ref: University/Postgraduate) (0.04) (0.04) (0.04) (0.04) (0.04)       d Secondary/Higher Secondary 0.184*** e 0.185*** 0.184*** 0.184*** 0.186***  (0.03) (0.03) (0.03) (0.03) (0.03)       e Above higher secondary 0.116*** c, d 0.117*** 0.117*** 0.117*** 0.118***  (0.02) (0.02) (0.02) (0.02) (0.02)       f Left (ref: Right) 0.351*** g 0.351*** 0.351*** 0.350*** 0.350***  (0.04) (0.04) (0.04) (0.03) (0.03)       g Centre 0.204*** f 0.204*** 0.203*** 0.202*** 0.202***  (0.04) (0.04) (0.04) (0.04) (0.04)       Female 0.106*** 0.107*** 0.106*** 0.106*** 0.106***  (0.01) (0.01) (0.01) (0.01) (0.01)       Age 0.003*** 0.003*** 0.003*** 0.003*** 0.003***  (0.00) (0.00) (0.00) (0.00) (0.00)       Married 0.002 0.002 0.002 0.002 0.002  (0.01) (0.01) (0.01) (0.01) (0.01)       120  Not religious (ref: Catholic) 0.014 0.015 0.014 0.013 0.014  (0.02) (0.02) (0.02) (0.02) (0.02)       Protestant -0.020 -0.020 -0.020 -0.022 -0.021  (0.02) (0.02) (0.02) (0.02) (0.02)       Other religion 0.021 0.022 0.022 0.022 0.023  (0.03) (0.03) (0.03) (0.03) (0.03) Societal-level      Gini 1.030 2.088* 1.031 1.012 1.851*  (0.75) (0.84) (0.75) (0.75) (0.83)       GDP -0.011 -0.011 -0.013 -0.010 -0.010  (0.03) (0.03) (0.03) (0.03) (0.03)       h Conservative (ref: Liberal) 0.406*** i 0.408*** 0.406*** 0.322** 0.380***  (0.09) (0.09) (0.09) (0.11) (0.11)       i Social Democratic -0.094 h, j, k -0.091 -0.094 -0.293** -0.217  (0.10) (0.10) (0.10) (0.11) (0.11)       j Former Communist 0.525*** i 0.525*** 0.525*** 0.555*** 0.603***  (0.10) (0.10) (0.10) (0.11) (0.11)       k Mediterranean 0.265* i 0.267* 0.264* 0.385** 0.394***  (0.12) (0.12) (0.12) (0.12) (0.12)       l 1992 (ref: 2009) -0.155* -0.155* -0.155* -0.155* -0.157*  (0.07) (0.07) (0.07) (0.07) (0.08)       m 1999 -0.082 -0.082 -0.082 -0.082 -0.083  (0.06) (0.06) (0.06) (0.06) (0.06) Interactions      121  Working Class*Gini (ref: Salariat)  -1.834***   -1.513***   (0.48)   (0.40)       Intermediate Class*Gini  -1.258**   -0.862   (0.46)   (0.47)       Working Class*GDP (ref: Salariat)   0.005      (0.02)         Intermediate Class*GDP   0.000      (0.01)         Working Class*Conservative    0.143* 0.034 (ref: Salariat Liberal)    (0.07) (0.07)       Working Class*Social Democratic    0.351*** 0.212**     (0.06) (0.07)       Working Class*Former Communist    -0.069 -0.159**     (0.06) (0.06)       Working Class*Mediterranean    -0.209*** -0.225***     (0.06) (0.06)       Intermediate Class*Conservative    0.104 0.044     (0.06) (0.07)       Intermediate Class*Social Democratic    0.259*** 0.181**     (0.06) (0.07)       Intermediate Class*Former Communist    -0.010 -0.061     (0.06) (0.06)       122  Intermediate Class*Mediterranean    -0.122* -0.131*     (0.06) (0.06)       Intercept 3.652*** 3.669*** 3.709*** 3.647*** 3.639***  (0.74) (0.74) (0.73) (0.74) (0.75) Random Effects      Country Wave .04*** .04*** .04*** .04*** .04***  (0.01) (0.01) (0.01) (0.01) (0.01) Class .01*** .01*** .01*** .01*** .00***  (0.00) (0.00) (0.00) (0.00) (0.00) Number of individuals 32627 32627 32627 32627 32627 Number of countries 24 24 24 24 24 Number of country year 53 53 53 53 53 AIC 81932 81917 81935 81901 81893 Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Notes: The letters next to the model 2 estimates indicates significant differences between the survey years, classes, education levels, and party affiliation types (*p < .05).    123  Table 4.3 Average marginal effects of the classes at various values of Gini. Results are from Model 2  Gini   Class .15 .2 .25 .30 .35 .40   a Working Class  4.160*** b, c  4.173*** b, c  4.186*** b, c  4.198*** b, c 4.211*** 4.224*** b Intermediate Class 4.036*** a, c  4.077*** a, c  4.119*** a, c  4.160*** a, c 4.202*** 4.243*** c Salariat Class 3.764*** a, b   3.869*** a, b  3.973*** a, b  4.078*** a, b 4.182***  4.286***  *** p < 0.001 Note: Letters next to average marginal effects indicate the predicted margins between classes that are significantly different across levels of Gini (*p < .05).    Table 4.4 Average marginal effects for classes across regime types. Results are from Model 4  Regime Type Class Liberal  Conservative Social Democratic Former Communist Mediterranean a Working 3.969*** b, c 4.433*** b, c 4.027*** b, c 4.455***  4.145*** c b Intermediate 3.908*** a 4.333*** a, c 3.875*** a, c 4.453*** 4.171*** c Salariat 3.840*** a, b 4.162*** a, b 3.547*** a, b 4.394*** 4.225*** a *** p < 0.001 Note: Letters next to average marginal effects indicate the predicted margins between classes that are significantly  different across regime types (*p < .05).                      124  Table 4.5 Average marginal effects for class across levels of Gini. Results are from Model 5  Gini   Class .15 .2 .25 .30 .35 .40   a Working Class  4.160*** b, c  4.176*** b, c  4.193*** b, c  4.210*** b, c 4.227*** c 4.244*** b Intermediate Class 4.014*** a, c  4.064*** a, c  4.113*** a, c  4.163*** a, c 4.212*** 4.261*** c Salariat Class 3.793*** a, b   3.885*** a, b  3.978*** a, b  4.070*** a, b 4.163*** a 4.255***  *** p < 0.001 Note: Letters next to average marginal effects indicate the predicted margins between classes that are significantly  different across Gini (*p < .05).     Table 4.6 Average marginal effects for class across regime type. Results are from Model 5  Regime Type Class Liberal  Conservative Social Democratic Former Communist Mediterranean a Working 3.997*** b, c 4.411*** b, c 3.992*** b, c 4.440***  4.166***  b Intermediate 3.909*** a, c 4.333*** a, c 3.873*** a, c 4.450*** 4.171*** c Salariat 3.807*** a, b 4.187*** a, b 3.590*** a, b 4.409*** 4.201***  *** p < 0.001 Note: Letters next to average marginal effects indicate the predicted margins between classes that are significantly  different across regime types (*p < .05).    125    Figure 4.3 Critical views towards inequality by Gini. Predictions are from Model 2 126  Figure 4.4 Critical perceptions of inequality according to regime type. Predictions are from Model 4 127  4.7. Discussion and Conclusion  This chapter has contributed to research on explaining country level variation in attitudes towards inequality. It improves upon the conceptual and empirical research on country level determinants of citizens’ critical views towards inequality, referring both towards the broader contextual factors and social class. I show that opinions towards inequality must be understood in terms of an individual’s placement within the social relations of production, but also with the broader contextual factors which impinge on individual subjectivities. The analyses reported suggest three overarching conclusions.  First, in isolation, macro-level performance of economic systems is not a significant contributor in determining citizens’ disapproval towards inequality. Second, citizens of former communist regimes tend to be more critical of inequality and show less class polarization than those in liberal or social democratic regimes. Third, the results indicate that, when it comes to critical attitudes towards inequality, people are not uniformly sensitive to different levels of the actual reality of inequality. Those who are located in the most advantaged class are more sensitive to higher levels of inequality than those in the most disadvantaged. In this way, objective inequality produces positive, but also heterogeneous consequences on critical perceptions of inequality, depending on an individual’s position with the social relations of production.  Motivated by Inglehart’s claim that economic development is a major factor in the shift to post-materialism values in modern societies, I investigated the relationship between economic growth and attitudes to income inequality. I found that economic growth, understood as GDP, does not affect how people view the economy. In particular, economic growth has little impact on the class differences in viewing inequality. A possible explanation as to why this is the case is that in “developed” nations the advantages of economic growth are of course not distributed equally across classes and the “success” of the economy may not register in the minds of citizens when they ponder the legitimacy of inequality in their society (Atkinson and Piketty 2007). The working classes gain less from economic growth and thus 128  their level of criticism of income inequality does not decrease as sharply as GDP increases, in comparison to the level of criticism of the salariat classes.  The findings that citizens in former communist regimes are more critical of inequality compared to market-oriented societies could reflect a lingering resentment with the shift to capitalism, which often brings with it substantial increases in material inequality. Perhaps, the “spectre of communism” does indeed exist in these nations and although they have adopted capitalist principles, the citizenry remain more cognizant and aware of the inequality that surrounds them. The results also reveal the interplay between class position and regime type in determining individual preferences towards inequality. Classes are more polarized in their attitudes towards inequality in social democratic regimes, whereas they are more similar in liberal regimes. This possibly speaks to the structure of class relations at the societal level shaping individual attitudes towards inequality. For instance, Sweden and the United States are in some respects opposite in terms of class formation, income inequality and welfare state programmes (Wright 1997).  The framing of the differing social policy programmes in social democratic and liberal regimes may increase the impact of class in the former while reducing it in the later. The strategies adopted by social democratic regimes often explicitly reference the notion of class. As Svallfors (2003) suggests, a key factor which heightens the class differences in social democratic regimes is the political articulation of class, which has often been employed by political parties in order to frame politics and construct policies. For example, in Sweden the structure of political parties has had a clear class base (Svallfors 2004). While the class foundations of political parties and policies may have waned in recent decades, they nevertheless may partly be responsible for the increased classed cleavages in disapproval with inequality among citizens of social democratic regimes. As a result, class has greater impact on perceptions in countries which adopt class discourse in social polices in comparison to countries which attempt to hide these. In Liberal regimes, certainly the United States, class, up until more recently, is a “zombie” concept and rarely discussed in the media or public discourse. In fact, there exists a multitude 129  of national mythologies in some liberal regimes which propagate the idea of classless societies (e.g. “American Dream” in the U.S., “Tall Poppy Syndrome” in the Antipodes, and so on).  I also entered into the debate between those that suggest that material inequality matters little in shaping peoples’ attitudes towards inequality and those that argue the opposite (Andersen and Fetner 2008; Carroll et al. 2011; McCall and Percheski 2010; McCall and Manza 2011). Like previous research, I also found no clear link between the actual level of inequality and discontent towards it (Larsen 2016; McCall and Kenworthy 2009). In the aggregate, people tend to reject the conditions in which they live regardless of the actual reality of inequality. On face value then, this would suggest that the economic reality does not affect economic evaluations. As I discuss further below, however, this is decidedly not the case when the economic position of the individual is taken into consideration.  I find that inequality and class-related aspects interact to influence perceptions of inequality. What is interesting is the pattern that exists between the classes. As inequality rises, it tends to be viewed as illegitimate across classes, but those who have the most to lose become increasingly critical, such that in contexts of high levels of inequality their condemnation surpasses those who are the most affected by it. I speculate that in more equal societies the advantaged classes are more likely to feel that inequality is justified as they may not see the effects of inequality that surround them. In unequal societies, on the other hand, the advantaged begin to be affected by inequality and increasingly come to view it negatively. Increased exposure to inequality may function to delegitimize acceptance of inequality. This may occur because in these contexts the dominant fractions of society are increasingly exposed to the multitude of social issues which coincide with heightened levels of inequality such as health problems (Carpiano, Link and Phelan 2008; Carroll et al. 2011; Elgar 2010; Wilkinson and Pickett 2010), increasing rates of criminal violence (Blau and Blau 1982), civic disengagement (Lancee and Van de Werfhorst 2012; Uslaner 2002), and reduced social trust (Fairbrother and Martin 2013; Hamamura 2012; Larsen 2013; Rözer and Volker 2016). 130  At lower levels of inequality, the intermediate and salariat classes are less likely to view inequality critically. Not surprisingly, the working classes are more likely to view inequality critically. As inequality increases, attitudes begin to converge as a greater portion of people adopt left-leaning values. However, the negative attitudes of the working classes towards inequality increase less sharply. This may suggest that in both unequal and more equal societies, the working classes are cognizant of the detriments of inequality and may recognize the benefits of state actions to reduce it. Attitudes towards inequality of the working classes do increase alongside the actual level of inequality but not as rapidly as the salariat. To explain why working-class disapproval towards inequality are gradual in these contexts, we must be mindful of their disadvantaged economic standing in relation to others. From an economic self-interest perspective, individuals in these positions are likely to be critical of inequality regardless of the level of inequality. Indeed, the pattern I find confirms that the working classes share a steady increase in their displeasure with inequality rather than a sharp surge characterized by their advantaged counterparts.  I conclude the chapter by discussing a possible policy implication of these results. People are increasingly critical of inequality as it rises, particularly among the advantaged. As inequality increases, a greater proportion of the population feel the negative effects of inequality. Higher levels of objective inequality in many societies brings the attitudes towards inequality of the various class fractions together. These shared attitudes towards inequality may lead to a greater desire for redistributive measures across classes which aims to reduce the unequal distribution of wealth. This may mean that, when inequality reaches a certain point, policy makers may feel the need to respond accordingly.     131  Chapter 5: Conclusion  The goal of this dissertation was to explore class views towards inequality across various countries. In doing so, I have offered new evidence that the positions people occupy in the social forces of production, coupled with their political preferences, influence attitudes towards economic inequality. I also demonstrate that country-level income inequality and regime type profoundly affect class attitudes towards inequality. People across classes become more critical of inequality as it increases because the effects of inequality in society becomes more evident. These results reveal that peoples’ preferences for inequality are strongly reflective of their class position, which bolsters an economic self-interests thesis. In the following sections I summarize the research contributions and outline implications of my research. I then conclude with a discussion of the research limitations and offer some suggestions for future work. 5.1. Re-animating Class  My dissertation offers compelling evidence that social class has an important role in shaping economic values. In Chapter two, I demonstrate that social class is associated with inequality views. I found that members of the most advantaged class are the least critical of inequality; whereas the working class are the most. This evidence calls into question theorists claiming that class is obsolete and that inequality views are unrelated. I also show the importance of separating social class and socioeconomic status when studying perceptions of inequality and reveal the impact of vote choice on orientations towards inequality.  Building from this research, chapter three assessed how social class and political ideology affect inequality views during a recession. I conclude that classes respond differently during recessions in terms of their views towards inequality. I found that TGR raised critical awareness of inequality among the “upper class” but not for the working class. I also offered evidence that these responses to TGR suggest class convergence in views towards inequality. This finding reveals that economic threats may have the capacity to stimulate class politics, which may mean that greater economic inequality could reduce class differences in viewing inequality critically.  132  In terms of politics, I show that after TGR the working class showed little political variation in their views towards inequality while their “upper class” showed more. This finding indicates that political persuasion plays a more important role in shaping inequality views among elites and this is enhanced after TGR. In times of heightened levels of economic turmoil, the most advantaged classes shift their attention to economic issues.  Building on this idea, chapter four assesses how individual and national economic conditions influence how people think about inequality. The main findings in this chapter relates to the “upper class”. At low levels of inequality, the class differences in viewing inequality critically are wide. I argue that under these conditions, the more advantaged classes see inequality as justified because they may not see the effects around them. However, at high levels of inequality, the advantaged begin to be affected more by inequality and come to view it negatively. This finding bolsters an economic self-interest perspective, that people are likely to support inequality when they feel that they are to benefit from it. When inequality begins to increase, a greater proportion of people are affected by it, such that even those who are in advantaged economic positions feel the effects of inequality.  5.2. Research Significance and Implications for Class Politics  This dissertation also holds further implications for class politics. Class politics are expected to shape how people view inequality and the extent to which they view this as just or unjust. Those favouring politics of the right are expected to view inequality less negatively than those on the left. As a result, the effect of actual inequality on opinions towards it should be muted among those on the right and more substantial among those on the left. Indeed, being located on the political right has been shown to be a powerful determinant of the degree to which the status quo and inequality are perceived as legitimate and justified (Anderson and Singer 2008). People view the performance of the macro economy in line with their political predispositions. People’s political dispositions should induce them to connect views towards the political system and inequality in different ways. My findings reveal, however, that people in less advantaged class positions share similar attitudes towards inequality regardless of their political 133  persuasion. This may translate to seeking policies to reduce inequality across partisan lines for those in working class positions. Moreover, it is possible that working class conservatives will rethink their views on other various social policies—generating more progressive attitudes towards redistribution for instance—which are commonplace in those who are more left-leaning.  Although inequality has the greatest material impact on the working classes, its consequences are becoming more acute also among the advantaged classes. Because of this, attitudes towards inequality become more critical across classes in contexts where its effects are most evident. If only the working class is critical of inequality and desire change it is easier for policy makers to ignore. However, if inequality continues to increase and it begins to affect those who hold power and political sway—such as those residing in dominant economic positions—politicians may feel more compelled to take action and intervene. As McCall (2011: 31) notes, “[t]he fact that responses to prompts about inequality cohere at all in such an unsettled environment . . . should be taken as a sign of their validity”. Indeed, it may be that this coming together of class politics may endure as inequality increases, which may in turn act as a catalyst for political change among the disenfranchised. If this pattern persists, growing inequality may raise class awareness which cuts across party lines and this may have implications for class politics. The connection between rising inequality and critical views towards inequality represents a possible site with which politicians could bring together people across party lines by speaking to their shared interests. Rather than obfuscating class vernacular in public discourse, coalition builders should target those in advantaged and disadvantaged positions by speaking directly to the benefits of progressive politics in terms of class. Rising inequality has ushered in concerns with inequality and this represents a site with which to create new alliances to bring together those class groups who possess divergent economic interests. A new political base premised on the social relations of production may rejuvenate social programs aimed at seriously reducing inequality and stimulating progressive politics which move beyond the neoliberal policies present in most Western democracies.  134  5.3. Future Research  This research is not without its limitations. Within sociology, the possible effects of individuals’ social positions on their perceptions of inequality has been extensively investigated (Hadler 2005; McCall and Percheski 2010; McCall 2013; Osberg and Smeeding 2006; Sealey and Andersen 2015; Svallfors 1997). Indeed, a large body of scholarship has explored how class-related aspects intersect with other dimensions of inequality, such as those organized around gender and race (Lamont 2000; Lareau 2003; Skeggs and Loveday 2012). That such effects are evident is unquestioned. While I did not seek to downplay these concerns, the focus of my dissertation was to raise a further issue that has so far received, in my opinion, remarkably little attention: that is the conceptualization and measurement of one’s social class and the implications that this has on perceptions of inequality. Evidently, the most notable limitation of my dissertation was that I was primarily focused on the role that social class and politics play in shaping public condemnation toward inequality. This focus ignores how other demographic characteristics such as gender, race, ethnicity and immigration status might affect inequality views.  While I included a “control” variable for respondent’s sex and found that women are in fact more critical of inequality than men, I did not parse out or attempt to elaborate on why this may be the case. Are women more likely to be critical of inequality and be more aware of its presence in society because they are often underpaid relative to men? Has this trend changed in recent decades because women are occupying more prominent roles in the labour market and increasingly advocating for equal pay? While some women may me more critical of inequality than men, it is possible that this could be conditioned by class. For instance, with many women increasingly occupying positions of privilege in the labour market, it would be interesting to explore whether residing in an advantaged class position moderates the gender effects on viewing inequality critically. Much more research is needed to examine this relationship.  In terms of race and ethnicity, it is well-established that immigrant populations face employment barriers (Berry, DeMeritt and Esarey 2010; Bonacich 1972; Fong and Wilkes 1999; Kaushal, Lu, Denier et al. 2016). New immigrants to Canada, for instance, predominately arrive through the point system 135  program. The point system designates points to prospective immigrants on the basis of language skills, education, work experience, age, employment prospects, and adaptability. Within the education category, more points are offered to those with higher levels of education. In this way, the point system works to select skilled immigrants with higher levels of education. While immigrants to Canada may be advantaged in terms of class position in country of origin, they may be disadvantaged in Canada in terms of employment prospects due to discrimination. It would be interesting to determine whether experiences of discrimination in the host country interacts with the social positioning among immigrant groups in determining their degree of resentment towards inequality.  I now provide some more general limitations. To begin, my research focus is entirely cross-national in nature. This focus affects the research questions that could be explored. Single country analyses are better able to provide a deeper and more historical account of how citizens in specific countries view inequality in their respective society. This research focus enables one to explore the possible national mythologies (American Dream, Tall Poppy Syndrome, etc.) that may persuade the citizenry to view inequality favourably. While I was able to explore how national-level economic and political factors shape inequality views, comparative work largely ignores potentially important country-level trends. Future research could analyze within country differences. This could be done by merging other ISSP datasets. For instance, the “Work Orientations”, “Role of Government”, “Social Inequality”, and “Social Networks” modules all include the variables necessary to create the measure of social class used in this analysis and could be merged to create the necessary data points to carry out an over-time analysis of single nations.  This merged data set could enable inequality views to be adapted to single country analyses. In fact, we know that income inequality and views toward it vary greatly within neighbourhoods in particular countries. For example, this is the case in Canada (Attewell, Kasinitz and Dunn 2010; Breau 2015; Fortin, Green, Lemieux et al. 2012; Sealey and Andersen 2015) the United States (Minkoff and 136  Lyons ; Norton and Ariely 2011; Osberg and Smeeding 2006) and parts of Europe (Oddsson 2010; Soroka and Wlezien 2014; Taylor-Gooby et al. 2019). I also need to acknowledge the selection of countries in this research. This study does not include several South American and African nations. Because their class structures are fundamentally different from those of more “developed” democracies, it could be that patterns of class polarization in inequality views is decidedly different. Indeed, scholars of some African societies argue that class formation is given “peripheral importance” in terms of shaping views towards society (Cohen 1972; Diamond 1987). Many African countries are dominated by ethnic, religious and regional politics. As Cohen notes, these important identities may distort class interests and thus may mute the influence of one’s economic standing on inequality views. Including such countries would allow for an exploration of particularly how class shapes perceptions of inequality in “developing” countries, and whether there are similar trajectories to more established economies.  Lastly, this research utilizes a single variable to determine critical inequality views. The ISSP social inequality module does include a battery of questions which also tap into various attitudes towards social inequality. In addition, in a number of the years, there are various questions regarding what people think people in certain occupations should earn and what people think said people should actually earn. Future work could employ more than one of these indicators of inequality views or create a multiple-item scale combining related aspects together. Such an analysis could reveal the possible relationship between class and inequality views towards various aspects in society beyond income differences.  5.4. Class as a Psychological Phenomenon The fact that [individuals] are not "class conscious" at all times and in all places does not mean that "there are no classes" . . . The economic and social facts are one thing. Psychological feelings may or may not be associated with them in rationally expected ways. Both are important, and if psychological  feelings . . . 137  do not correspond  to economic or occupational class, we must try to find out why, rather than throw out the economic baby with the psychological bath, and so fail to understand how either fits into the national tub (Mills 1970: 317).  In this dissertation, I have empirically established that by knowing the class position of an individual, we can predict their views towards inequality. As we know from the late neo-Marxist Wright (1996: 700), if ‘class analysis [is] to constitute a research program worth pursuing, it is sufficient that it identify important causal mechanisms; it is not necessary that class be the most important or fundamental determinant of social phenomena’. As socio-relational categories, classes induce subjectively salient experiences that structure the views used by individuals to legitimize inequality within stratified societies. I have empirically documented how subjective evaluations of inequality derive from objective class attributes. Taken together, this dissertation has contributed to reformulating the subjective implications of class by documenting that social class is an important component of how individuals view inequality in their society.  We are aware that the social world is not only subjectively constituted, but also objectively in terms of class and a variety of factors such as age, sex, religion, race, etc. The aim of this dissertation was not to completely explain what motivates individuals to view inequality critically; rather the desire was to empirically determine how class position, alongside aspects of socioeconomic status and politics, might exert an influence on these views. As Marxist classes are less easily operationalized for quantitative analysis, I turned to a Weberian measurement of class. Such a measurement of class has revealed the underpinnings of a Marxist theory, which purports that class, especially, impacts on subjectivities. I have documented that social class provides an incomplete yet moderately powerful explanation of why individuals across various regimes and types of economies view inequality negatively. This goes against those who claim that class-politics are endangered and dying (Beck and Beck-Gernsheim 2002). I empirically documented that not only are classed perceptions of material differences alive, they have 138  endured. Individuals do respond to inequality in a way that relates to their social position in the class structure.  Awareness of inequality is also increasing alongside global developments (Wilkinson and Pickett 2018). Class convergence in harshly judging inequality is occurring possibly because citizenry are being crushed by crippling austerity measures employed by neoliberal regimes in order to remain competitive globally. Citizens in many countries may be (re)entering a period of intense class awareness. Shared views towards inequality represents the first stage in a possible response which may lead to social action leading to a state response (McCall and Kenworthy 2009). As economic inequalities become more salient, people may become more aware of the causes of inequality and increasingly challenge them (Kluegel and Smith 1986; McCall and Kenworthy 2009). In this way, how individuals view economic disparities and whether they regard them as fair or unfair are important with respect to the fundamental stability of modern democracies (Sachweh 2012; Wilkinson and Pickett 2018). It is evident that more work is needed to consider the interplay between class, inequality views and political engagement. My results show that classes appear to be coalescing in terms of their displeasure with inequality. I leave it up to other scholars to uncover whether the proletariat will crystallise their shared views and ascend ‘the steps that [have] now become objectively possible’ towards true class consciousness (Lukács 1968:75). What is now required is a political leader bold enough to explicitly use the discourse of class vernacular in order to bring this group together.    139  Bibliography Allen, Kieran. 2004. Max Weber: A Critical Introduction. London: Pluto Press. Alvaredo, Facundo, Anthony B. Atkinson, Thomas Piketty and Emmanuel Saez. 2013. "The Top 1 Percent in International and Historical Perspective." The Journal of Economic Perspectives 27(3):3-20. Andersen, Robert and Anthony Heath. 2002. "Class Matters: The Persisting Effects of Contextual Social Class on Individual Voting in Britain, 1964-97." European Sociological Review 18(2):125-38. Andersen, Robert, Min Yang and Anthony F. Heath. 2006. "Class Politics and Political Context in Britain, 1964-1997: Have Voters Become More Individualized?". European Sociological Review 22(2):215-28. Andersen, Robert and Tina Fetner. 2008. "Economic Inequality and Intolerance: Attitudes toward Homosexuality in 35 Democracies." American Journal of Political Science 52(4):942-58. Andersen, Robert and Josh Curtis. 2012. "The Polarizing Effect of Economic Inequality on Class Identification: Evidence from 44 Countries." Research in Social Stratification and Mobility 30(1):129-41. Andersen, Robert and Meir Yaish. 2012. "Public Opinion on Income Inequality in 20 Democracies: The Enduring Impact of Social Class and Economic Inequality." Retrieved May 2, 2016 (http://www.uva-aias.net/uploaded_files/publications/DP48-Andersen,Yaish.pdf). Anderson, C. and J. Hecht. 2014. "Crisis of Confidence? : The Dynamics of Economic Opinions During the Great Recession." Pp. 40-71 in Mass Politics in Tough Times: Opinions, Votes and Protest in the Great Recession. Oxford: Oxford University Press. Anderson, Christopher J. and Matthew M. Singer. 2008. "The Sensitive Left and the Impervious Right:Multilevel Models and the Politics of Inequality, Ideology, and Legitimacy in Europe." Comparative Political Studies 41(4-5):564-99. Anthony, B. Atkinson. 2008. The Changing Distribution of Earnings in Oecd Countries. Oxford: Oxford University Press. Arts, Wil and Mérove Gijsberts. 1998. "After the Velvet Revolutions: Altered Life-Chances, Fragile Legitimacy, and Split-Consciousness in Post-Communist Eastern Europe." Social Justice Research 11(2):143-71. Arts, Wil and John Gelissen. 2002. "Three Worlds of Welfare Capitalism or More? A State-of-the-Art Report." Journal of European Social Policy 12(2):137-58. Atkinson, Anthony B. and Thomas Piketty. 2007. Top Incomes over the Twentieth Century: A Contrast between European and English Countries. Oxford: Oxford University Press. Atkinson, Anthony B., Thomas Piketty and Emmanuel Saez. 2011. "Top Incomes in the Long Run of History." Journal of Economic Literature 49(1):3-71. Atkinson, Will. 2007. "Beck, Individualization and the Death of Class: A Critique." The British Journal of Sociology 58(3):349-66. Atkinson, Will. 2010. "The Myth of the Reflexive Worker: Class and Work Histories in Neo-Liberal Times." Work, Employment & Society 24(3):413-29. Attewell, Paul, Philip Kasinitz and Kathleen Dunn. 2010. "Black Canadians and Black Americans: Racial Income Inequality in Comparative Perspective." Ethnic and Racial Studies 33(3):473-95. Baltodano, Marta. 2012. "Neoliberalism and the Demise of Public Education: The Corporatization of Schools of Education." International Journal of Qualitative Studies in Education 25(4):487-507. doi: 10.1080/09518398.2012.673025. Bartels, Larry M. 2005. "Homer Gets a Tax Cut: Inequality and Public Policy in the American Mind." Perspectives on Politics 3(1):15-31. Bartels, Larry M. 2013. "Political Effects of the Great Recession." The Annals of the American Academy of Political and Social Science 650:47-75. Bauman, Zygmunt. 1982. Memories of Class: The Pre-History and after-Life of Class. London: Routledge and Keagan Paul. 140  Beck, Ulrich and Elisabeth Beck-Gernsheim. 2002. Individualization: Institutionalized Individualism and Its Social and Political Consequences. London: Sage. Bell, Andrew and Kelvyn Jones. 2014. "Explaining Fixed Effects: Random Effects Modeling of Time-Series Cross-Sectional and Panel Data." Political Science Research and Methods 3(1):133-53. Bengtsson, Mattias, Tomas Berglund and Maria Oskarson. 2013. "Class and Ideological Orientations Revisited: An Exploration of Class-Based Mechanisms." The British Journal of Sociology 64(4):691-716. Bergman, Manfred Max and Dominique Joye. 2005. "Comparing Social Stratification Schemata: Camsis, Csp-Ch, Goldthorpe, Isco-88, Treiman, and Wright." Cambridge Studies in Social Research 10 [online]. Berry, William D., Jacqueline H. R. DeMeritt and Justin Esarey. 2010. "Testing for Interaction in Binary Logit and Probit Models: Is a Product Term Essential?". American Journal of Political Science 54(1):248-66. doi: 10.2307/20647982. Blau, Judith R. and Peter M. Blau. 1982. "The Cost of Inequality: Metropolitan Structure and Violent Crime." American Sociological Review 47(1):114-29. Bonacich, Edna. 1972. "A Theory of Ethnic Antagonism: The Split Labor Market." American Sociological Review 37(5):547-59. Bottero, Wendy. 2010. "Intersubjectivity and Bourdieusian Approaches to ‘Identity’." Cultural Sociology 4(1):3-22. Breau, Sébastien. 2015. "Rising Inequality in Canada: A Regional Perspective." Applied Geography 61:58-69. Breznau, Nate. 2010. "Economic Equality and Social Welfare: Policy Preferences in Five Nations." International Journal of Public Opinion Research 22(4):458-84. Brooks, Clem and Jeff Manza. 1997a. "The Social and Ideological Bases of Middle-Class Political Realignment in the United States, 1972 to 1992." American Sociological Review 62(2):191-208. Brooks, Clem and Jeff Manza. 1997b. "Class Politics and Political Change in the United States, 1952-1992." Social Forces 76(2):379-408. Brooks, Clem and Stefan Svallfors. 2010. "Why Does Class Matter? Policy Attitudes, Mechanisms, and the Case of the Nordic Countries." Research in Social Stratification and Mobility 28(2):199-213. Brooks, Clem and Jeff Manza. 2013. "A Broken Public? Americans' Responses to the Great Recession." American Sociological Review 78(5):727-48. Bryan, Mark L. and Stephen P. Jenkins. 2016. "Multilevel Modelling of Country Effects: A Cautionary Tale." European Sociological Review 32(1):3-22. Caínzos, Miguel and Carmen Voces. 2010. "Class Inequalities in Political Participation and the ‘Death of Class’ Debate." International Sociology 25(3):383-418. Carpiano, Richard M., Bruce G. Link and Jo C. Phelan. 2008. "Social Inequality and Health: Future Directions for the Fundamental Cause Explanation." Pp. 232-63 in Social Class: How Does It Work?, edited by A. Lareau and D. Conley. New York: Russell Sage Foundation. Carroll, Penelope, Sally Casswell, John Huakau, Philippa Howden-Chapman and Paul Perry. 2011. "The Widening Gap: Perceptions of Poverty and Income Inequalities and Implications for Health and Social Outcomes." Social Policy Journal of New Zealand (37):111-22. Cavaillé, Charlotte and Kris-Stella Trump. 2015. "The Two Facets of Social Policy Preferences." The Journal of Politics 77(1):146-60. Chan, Tak and John Goldthorpe. 2005. "The Social Stratification of Theatre, Dance and Cinema Attendance." Cultural Trends 14(3):193-212. Chan, Tak Wing and John H. Goldthorpe. 2007a. "Class and Status: The Conceptual Distinction and Its Empirical Relevance." American Sociological Review 72(4):512-32. Chan, Tak Wing and John H. Goldthorpe. 2007b. "Social Status and Newspaper Readership." American Journal of Sociology 112(4):1095-134. Chan, Tak Wing and John H. Goldthorpe. 2007c. "Social Stratification and Cultural Consumption: Music in England." European Sociological Review 23(1):1-19. 141  Chan, Tak Wing, Gunn Elisabeth Birkelund, Arne Kristian Aas and Øyvind Wiborg. 2011. "Social Status in Norway." European Sociological Review 27(4):451-68. Chandy, Laurence and Brina Seidl. 2018. "How Much Do We Really Know About Inequality within Countries around the World? Adjusting Gini Coefficients for Missing Top Incomes." in Brookings, Vol. Retrieved January 22, 2018 (https://www.brookings.edu/opinions/how-much-do-we-really-know-about-inequality-within-countries-around-the-world/). Clark, Terry Nichols and Seymour Martin Lipset. 1991. "Are Social Classes Dying?". International Sociology 6(4):397-410. Cohen, Robin. 1972. "Class in Africa : Analytical Problems and Perspectives." The Socialist Register 9:231-55. Cranford, Cynthia J. and Leah F. Vosko. 2006. "Conceptualizing Precarious Employment: Mapping Wage Work across Social Location and Occupational Context." in Precarious Employment: Understanding Labour Market Insecurity in Canada, edited by L. F. Vosko. Montreal: McGill-Queen's University Press. Curtis, Josh and Robert Andersen. 2015. "How Social Class Shapes Attitudes on Economic Inequality: The Competing Forces of Self-Interest and Legitimation." International Review of Social Research 5(1):4-19. Curtis, Josh. 2016. "Social Mobility and Class Identity: The Role of Economic Conditions in 33 Societies, 1999–2009." European Sociological Review 32(1):108-21. Dallinger, Ursula. 2010. "Public Support for Redistribution: What Explains Cross-National Differences?". Journal of European Social Policy 20(4):333-49. Davies, Bronwyn and Peter Bansel. 2007. "Neoliberalism and Education." International Journal of Qualitative Studies in Education 20(3):247-59. De Vries, Catherine E., Sara B. Hobolt and James Tilley. 2018. "Facing up to the Facts: What Causes Economic Perceptions?". Electoral Studies 51:115-22. Diamond, Larry. 1987. "Class Formation in the Swollen African State." The Journal of Modern African Studies 25(4):567-96. Dimick, Matthew, David Rueda and Daniel Stegmueller. 2017. "The Altruistic Rich? Inequality and Other-Regarding Preferences for Redistribution." Quarterly Journal of Political Science 11(4):385-439. Dimick, Matthew, David Rueda and Daniel Stegmueller. 2018. "Models of Other-Regarding Preferences, Inequality, and Redistribution." Annual Review of Political Science 21(1):441-60. Dodson, Kyle. 2017. "Economic Change and Class Conflict over Tax Attitudes: Evidence from Nine Advanced Capitalist Democracies." Social Forces 95(4):1509-38. Dodson, Lisa. 2007. "Wage-Poor Mothers and Moral Economy." Social Politics: International Studies in Gender, State & Society 14(2):258-80. Duch, Raymond M. and Iñaki Sagarzazu. 2014. "Crisis Perceptions and Economic Voting among the Rich and the Poor: The United Kingdom and Germany." Pp. 224-65 in Mass Politics in Tough Times: Opinions, Votes and Protest in the Great Recession, edited by L. Bartels and N. G. Bermeo. Oxford: Oxford University Press. Durante, Federica and Susan T. Fiske. 2017. "How Social-Class Stereotypes Maintain Inequality." Current Opinion in Psychology 18(Supplement C):43-48. Durkheim, Emile. [1893] 1984. The Division of Labour in Society. Hampshire: Macmillan. Durkheim, Emile. [1897] 2002. Suicide : A Study in Sociology. London: Routledge. Edlund, Jonas. 2003. "The Influence of the Class Situations of Husbands and Wives on Class Identity, Party Preference and Attitudes Towards Redistribution: Sweden, Germany and the United States." Acta Sociologica 46(3):195-214. Edlund, Jonas and Arvid Lindh. 2015. "The Democratic Class Struggle Revisited:The Welfare State, Social Cohesion and Political Conflict." Acta Sociologica 58(4):311-28. Elgar, Frank J. 2010. "Income Inequality, Trust, and Population Health in 33 Countries." American Journal of Public Health 100(11):2311-15. 142  Erikson, Robert and John H. Goldthorpe. 1992. The Constant Flux: A Study of Class Mobility in Industrial Societies. New York: Clarendon Press. Esping-Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism. Princeton: Princeton University Press. Etzioni, Amitai. 2014. "Politics and Culture in an Age of Austerity." International Journal of Politics, Culture, and Society 27(4):389-407. Evans, Geoffrey. 1992. "Testing the Validity of the Goldthorpe Class Schema." European Sociological Review 8(3):211-32. Evans, Geoffrey. 1996. "Putting Men and Women into Classes: An Assessment of the Cross-Sex Validity of the Goldthorpe Class Schema." Sociology 30(2):209-34. Evans, Geoffrey and Colin Mills. 1998. "Identifying Class Structure: A Latent Class Analysis of the Criterion-Related and Construct Validity of the Goldthorpe Class Schema." European Sociological Review 14(1):87-106. Evans, Geoffrey and Colin Mills. 2000. "In Search of the Wage-Labour/Service Contract: New Evidence on the Validity of the Goldthorpe Class Schema." The British Journal of Sociology 51(4):641-61. Evans, Geoffrey. 2002. "Class Voting: From Premature Obituary to Reasoned Appraisal " Pp. 1-20 in The End of Class Politics? Class Voting in Comparative Context, edited by G. Evans. Oxford, UK: Oxford University Press. Evans, Geoffrey and Robert Andersen. 2006. "The Political Conditioning of Economic Perceptions." The Journal of Politics 68(1):194-207. Evans, Geoffrey and Mark Pickup. 2010. "Reversing the Causal Arrow: The Political Conditioning of Economic Perceptions in the 2000-2004 U.S. Presidential Election Cycle." The Journal of Politics 72(4):1236-51. Evans, Geoffrey and James Tilley. 2012a. "The Depoliticization of Inequality and Redistribution: Explaining the Decline of Class Voting." The Journal of Politics 74(4):963-76. Evans, Geoffrey and James Tilley. 2012b. "How Parties Shape Class Politics: Explaining the Decline of the Class Basis of Party Support." British Journal of Political Science 42(01):137-61. Evans, Geoffrey and James Tilley. 2017. The New Politics of Class: The Political Exclusion of the British Working Class. Oxford: Oxford University Press. Evans, M. and Jonathan Kelley. 2018. "Strong Welfare States Do Not Intensify Public Support for Income Redistribution, but Even Reduce It among the Prosperous: A Multilevel Analysis of Public Opinion in 30 Countries." Societies 8(4):105. Evans, M. D. R. and Jonathan Kelley. 2017. "Communism, Capitalism, and Images of Class: Effects of Reference Groups, Reality, and Regime in 43 Nations and 110,000 Individuals, 1987-2009." Cross-Cultural Research 51(4):315-59. Fairbrother, Malcolm and Isaac W. Martin. 2013. "Does Inequality Erode Social Trust? Results from Multilevel Models of Us States and Counties." Social Science Research 42(2):347-60. Fong, Eric and Rima Wilkes. 1999. "The Spatial Assimilation Model Reexamined: An Assessment by Canadian Data." International Migration Review 33(3):594-620. Fortin, Nicole, David A. Green, Thomas Lemieux, Kevin Milligan and W. Craig Riddell. 2012. "Canadian Inequality: Recent Developments and Policy Options." Canadian Public Policy / Analyse de Politiques 38(2):121-45. Franko, William W. 2017. "Understanding Public Perceptions of Growing Economic Inequality." State Politics & Policy Quarterly 17(3):319-48. Gabennesch, Howard. 1972. "Authoritarianism as World View." American Journal of Sociology 77(5):857-75. Galbraith, James K. 2014. Inequality and Instability : A Study of the World Economy Just before the Great Crisis. Oxford: Oxford University Press. Gane, Nicholas. 2012. Max Weber and Contemporary Capitalism. Hampshire: Palgrave Macmillan. Gelissen, J. 2000. "Popular Support for Institutionalised Solidarity: A Comparison between European Welfare States." International Journal of Social Welfare 9(4):285-300. 143  German, Danielle and Carl A. Latkin. 2012. "Social Stability and Health: Exploring Multidimensional Social Disadvantage." Journal of urban health : bulletin of the New York Academy of Medicine 89(1):19-35. Giddens, Anthony. 1991. Modernity and Self-Identity: Self and Society in the Late Modern Age. Cambridge: Polity Press. Gijsberts, Mérove. 2002. "The Legitimation of Income Inequality in State-Socialist and Market Societies." Acta Sociologica 45(4):269-85. Giuliani, Marco and Sergio Alberto Massari. 2017. "The Economic Vote at the Party Level: Electoral Behaviour During the Great Recession." Party Politics:1-13. Global News. 2019, "Canada Election: The 2019 Results by the Numbers". Retrieved 2019. 11, 29,  (https://globalnews.ca/news/6066524/canada-election-the-2019-results-by-the-numbers/). Goldthorpe, John. 2008. "Two Oppositions in Studies of Class: A Reflection." Pp. 350-53 in Social Class: How Does It Work?, edited by A. Lareau and D. Conley. New York: Russell Sage Foundation. Goldthorpe, John H., David Lockwood, Frank Bechhofer and Jennifer Platt. 1970. The Affluent Worker: Industrial Attitudes and Behaviour. Cambridge: Cambridge University Press. Goldthorpe, John H. 1996. "Class Analysis and the Reorientation of Class Theory: The Case of Persisting Differentials in Educational Attainment." The British Journal of Sociology 47(3):481-505. Goldthorpe, John H. 2002. "Occupational Sociology, Yes: Class Analysis, No: Comment on Grusky and Weeden's "Research Agenda"." Acta Sociologica 45(3):211-17. Goldthorpe, John H. 2010. "Analysing Social Inequality: A Critique of Two Recent Contributions from Economics and Epidemiology." European Sociological Review 26(6):731-44. Goldthorpe, John. H and Abigail McKnight. 2006. "The Economic Basis of Social Class." Pp. 109-36 in Mobility and Inequality: Frontiers of Research in Sociology and Economics, edited by S. L. Morgan, D. B. Grusky and G. S. Fields. Stanford: Stanford University Press. Gonthier, Frederic. 2017. "Parallel Publics? Support for Income Redistribution in Times of Economic Crisis." European Journal of Political Research 56(1):92-114. Haddon, Edward. 2015. "Class Identification in New Zealand: An Analysis of the Relationship between Class Position and Subjective Social Location." Journal of Sociology 51(3):737-54. Hadler, Markus. 2005. "Why Do People Accept Different Income Ratios? A Multi-Level Comparison of Thirty Countries." Acta Sociologica 48(2):131-54. Hamamura, Takeshi. 2012. "Social Class Predicts Generalized Trust but Only in Wealthy Societies." Journal of Cross-Cultural Psychology 43(3):498-509. Harwell, Michael R. and Guido G. Gatti. 2001. "Rescaling Ordinal Data to Interval Data in Educational Research." Review of Educational Research 71(1):105-31. Heisig, Jan Paul and Merlin Schaeffer. 2019. "Why You Should Always Include a Random Slope for the Lower-Level Variable Involved in a Cross-Level Interaction." European Sociological Review 35(2):258-79. Hicks, Timothy, Alan M. Jacobs and J. Scott Matthews. 2016. "Inequality and Electoral Accountability: Class-Biased Economic Voting in Comparative Perspective." The Journal of Politics 78(4):1076-93. Hout, Michael, Asaf Levanon and Esra Burak. 2011. "Job Loss and Unemployment." Pp. 59-81, edited by D. B. Grusky, B. Western and C. Wimer. New York: Russell Sage Foundation. Hoynes, Hilary W., Douglas L. Miller and Jessamyn Schaller. 2012. "Who Suffers During Recessions?". NBER Working Paper No. 17951 Retrieved November 11, 2019 (https://www.nber.org/papers/w17951.pdf). Hyman, Herber Hiram and Charles Robert Wright. 1979. Education's Lasting Influence on Values. Chicago: University of Chicago Press,. Ignácz, Zsófia. 2018. "The Remains of the Socialist Legacy: The Influence of Socialist Socialization on Attitudes toward Income Inequality." Societies 8(3):62. Im, Dong-Kyun. 2014. "The Legitimation of Inequality: Psychosocial Dispositions, Education, and Attitudes toward Income Inequality in China." Sociological Perspectives 57(4):506-25. 144  Inglehart, Ronald. 1971. "The Silent Revolution in Europe: Intergenerational Change in Post-Industrial Societies." The American Political Science Review 65(4):991-1017. Inglehart, Ronald and Jacques-René Rabier. 1986. "Political Realignment in Advanced Industrial Society: From Class-Based Politics to Quality-of-Life Politics." Government and Opposition 21(4):456-79. Inglehart, Ronald and Scott C. Flanagan. 1987. "Value Change in Industrial Societies." The American Political Science Review 81(4):1289-319. Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Society. Princeton: Princeton University Press. Inglehart, Ronald and Paul R. Abramson. 1994. "Economic Security and Value Change." The American Political Science Review 88(2):336-54. Inglehart, Ronald and Paul R. Abramson. 1999. "Measuring Postmaterialism." The American Political Science Review 93(3):665-77. Inglehart, Ronald and Christian Welzel. 2005. Modernization, Cultural Change, and Democracy: The Human Development Sequence Cambridge University Press: Cambridge. Inglehart, Ronald and Pippa Norris. 2017. "Trump and the Populist Authoritarian Parties: The Silent Revolution in Reverse." Perspectives on Politics 15(2):443-54. Irwin, Sarah. 2018. "Lay Perceptions of Inequality and Social Structure." Sociology 52(2):211-27. Iversen, Torben and David Soskice. 2015. "Information, Inequality, and Mass Polarization: Ideology in Advanced Democracies." Comparative Political Studies 48(13):1781-813. Jackman, Mary R. and Michael J. Muha. 1984. "Education and Intergroup Attitudes: Moral Enlightenment, Superficial Democratic Commitment, or Ideological Refinement?". American Sociological Review 49(6):751-69. Janmaat, Jan Germen. 2013. "Subjective Inequality: A Review of International Comparative Studies on People’s Views About Inequality." European Journal of Sociology 54(03):357-89. Jepperson, Ronald. 1991. "Institutions, Institutional Effects, and Institutionalism " Pp. 143-63 in The New Institutionalism in Organizational Analysis, edited by W. W. Powell and P. J. DiMaggio. Chicago: University of Chicago Press. Jetten, Jolanda, Frank Mols, Nikita Healy and Russell Spears. 2017. "“Fear of Falling”: Economic Instability Enhances Collective Angst among Societies’ Wealthy Class." Journal of Social Issues 73(1):61-79. Johnson, David Richard and James C. Creech. 1983. "Ordinal Measures in Multiple Indicator Models: A Simulation Study of Categorization Error." American Sociological Review 48(3):398-407. Jordan, Jason. 2016. "Religion and Inequality: The Lasting Impact of Religious Traditions and Institutions on Welfare State Development." European Political Science Review 8(1):25-48. Kalleberg, Arne L. 2000. "Nonstandard Employment Relations: Part-Time, Temporary and Contract Work." Annual Review of Sociology 26:341-65. Kaushal, Neeraj, Yao Lu, Nicole Denier, Julia Shu-huah Wang and Stephen J. Trejo. 2016. "Immigrant Employment and Earnings Growth in Canada and the USA: Evidence from Longitudinal Data." Journal of Population Economics 29(4):1249-77. Kelley, Jonathan and M. D. R. Evans. 1993. "The Legitimation of Inequality: Occupational Earnings in Nine Nations." American Journal of Sociology 99(1):75-125. Kelley, Jonathan and M. D. R. Evans. 1995. "Class and Class Conflict in Six Western Nations." American Sociological Review 60(2):157-78. Kelley, Jonathan and Krzysztof Zagorski. 2004. "Economic Change and the Legitimation of Inequality: The Transition from Socialism to the Free Market in Central-East Europe." Research in Social Stratification and Mobility 22:319-64. Kelly, Nathan J. and Peter K. Enns. 2010. "Inequality and the Dynamics of Public Opinion: The Self-Reinforcing Link between Economic Inequality and Mass Preferences." American Journal of Political Science 54(4):855-70. 145  Kenworthy, Lane and Leslie McCall. 2008. "Inequality, Public Opinion and Redistribution." Socio-Economic Review 6(1):35-68. Kenworthy, Lane and Lindsay A. Owens. 2011. "The Surprisingly  Weak Effect of Recessions  on Public Opinion." in The Great Recession, edited by D. B. Grusky, B. Western and C. Wimer. New York: Russell Sage Foundation. Kenworthy, Lane and Lindsay A. Owens. 2012. "Political Attitudes, Public Opinion, and the Great Recession." Retrieved May 8, 2018 (https://web.stanford.edu/group/recessiontrends/cgi-bin/web/sites/all/themes/barron/pdf/PublicOpinion_fact_sheet.pdf). Kingston, Paul W. 2000. The Classless Society. Stanford: Stanford University Press. Kluegel, James R and Eliot R Smith. 1986. Beliefs About Inequality : Americans' Views of What Is and What Ought to Be. New York: de Gruyter. Koçer, Rüya Gökhan and Herman G. van de Werfhorst. 2012. "Does Education Affect Opinions on Economic Inequality? A Joint Mean and Dispersion Analysis." Acta Sociologica 55(3):251-72. Kulin, Joakim and Stefan Svallfors. 2013. "Class, Values, and Attitudes Towards Redistribution: A European Comparison." European Sociological Review 29(2):155-67. Lamont, Michèle. 2000. The Dignity of Working Men: Morality and the Boundaries of Race, Class, and Immigration. New York: Russel Sage Foundation. Lancee, Bram and Herman G. Van de Werfhorst. 2012. "Income Inequality and Participation: A Comparison of 24 European Countries." Social Science Research 41(5):1166-78. Lareau, Annette. 2003. Unequal Childhoods: Class, Race and Family Life. Berkeley: University  of California Press. Larsen, Christian. 2013. "Broken Societies: Inequality, Cohesion and the Middle-Class Dream." Juncture 20(3):193-99. Larsen, Christian Albrekt. 2008. "The Institutional Logic of Welfare Attitudes: How Welfare Regimes Influence Public Support." Comparative Political Studies 41(2):145-68. Larsen, Christian Albrekt. 2016. "How Three Narratives of Modernity Justify Economic Inequality." Acta Sociologica 59(2):93-111. Laurin, Kristin, Danielle Gaucher and Aaron Kay. 2013. "Stability and the Justification of Social Inequality." European Journal of Social Psychology 43(4):246-54. Lukács, Georg. 1968. History and Class Consciousness: Studies in Marxist Dialectics. Cambridge: MIT Press. Luttig, Matthew. 2013. "The Structure of Inequality and Americans’ Attitudes toward Redistribution." Public Opinion Quarterly 77(3):811-21. Margalit, Yotam M. 2013. "Explaining Social Policy Preferences: Evidence from the Great Recession." The American Political Science Review 107(1):80-103. Marks, Gary N. 2005. "Cross-National Differences and Accounting for Social Class Inequalities in Education." International Sociology 20(4):483-505. Marshall, Gordon, Howard Newby, David Rose and Carolyn Vogler. 1988. Social Class in Modern Britain. London: Hutchinson. Marx, Karl. [1847] 2012. "Wage-Labour and Capital." Pp. 182-89 in Classical Sociological Theory, edited by C. Calhoun, J. Gerteis, J. Moody, S. Pfaff and I. Virk. West Sussex: Wiley-Blackwell. Marx, Karl. [1848] 2012. "Manifesto of the Communist Party." Pp. 156-71 in Classical Sociological Theory, edited by C. Calhoun, J. Gerteis, J. Moody, S. Pfaff and I. Virk. West Sussex: Wiley-Blackwell. Marx, Karl. [1867] 2012. "Classes." Pp. 190-91 in Classical Sociological Theory, edited by C. Calhoun, J. Gerteis, J. Moody, S. Pfaff and I. Virk. West Sussex: Wiley-Blackwell. Mayer, Nonna. 2014. "The Electoral Impact of the Crisis on the French Working Class: More to the Right?" Pp. 266-96 in Mass Politics in Tough Times: Opinions, Votes and Protest in the Great Recession, edited by L. Bartels and N. G. Bermeo. Oxford: Oxford University Press. McCall, Leslie and Lane Kenworthy. 2009. "Americans' Social Policy Preferences in the Era of Rising Inequality." Perspectives on Politics 7(3):459-84. 146  McCall, Leslie and Christine Percheski. 2010. "Income Inequality: New Trends and Research Directions." Annual Review of Sociology 36(1):329-47. McCall, Leslie and Jeff Manza. 2011. "Class Differences in Social and Political Attitudes." Pp. 552-70 in The Oxford Handbook of American Public Opinion and the Media, edited by George C. Edwards III, Lawrence R. Jacobs and Robert Y. Shapiro. New York: Oxford University Press. McCall, Leslie. 2013. The Undeserving Rich: American Beliefs About Inequality, Opportunity, and Redistribution. Cambridge: Cambridge University Press. McCall, Leslie. 2016. "Political and Policy Responses to Problems of Inequality and Opportunity: Past, Present, and Future." Pp. 415-42 in The Dynamics of Opportunity in America: Evidence and Perspectives, edited by I. Kirsch and H. Braun. Cham: Springer International Publishing. Meltzer, Allan H. and Scott F. Richard. 1981. "A Rational Theory of the Size of Government." Journal of Political Economy 89(5):914-27. Milanovic, Branko. 2013. Inequality Possibility Frontier : Extensions and New Applications. Washington, DC: World Bank. Mills, C. Wright. 1970. "The Sociology of Stratification." Pp. 305-23 in Power, Politics and People, edited by I. L. Horowitz. London: Oxford University Press. Milojev, Petar, Lara Greaves, Danny Osborne and Chris G. Sibley. 2015. "Stability and Change in Political Conservatism Following the Global Financial Crisis." Personality and Social Psychology Bulletin 41(1):127-39. Minkoff, Scott L. and Jeffrey Lyons. 2019. "Living with Inequality: Neighborhood Income Diversity and Perceptions of the Income Gap." American Politics Research 47(2). Modrek, Sepideh and Mark R Cullen. 2013. "Job Insecurity During Recessions: Effects on Survivors ’ Work Stress." BMC Public Health 13:929-40. Mols, Frank and Jolanda Jetten. 2017. The Wealth Paradox: Economic Prosperity and the Hardening of Attitudes. Cambridge: Cambridge University Press. Moya, Miguel and Susan T. Fiske. 2017. "The Social Psychology of the Great Recession and Social Class Divides." Journal of Social Issues 73(1):8-22. Mustillo, Sarah. 2012. "The Effects of Auxiliary Variables on Coefficient Bias and Efficiency in Multiple Imputation." Sociological Methods & Research 41(2):335-61. Newman, Benjamin J., Sono Shah and Erinn Lauterbach. 2018. "Who Sees an Hourglass? Assessing Citizens’ Perception of Local Economic Inequality." Research & Politics 5(3). Niedzwiedz, Claire L. and Mor Kandlik-Eltanani. 2014. "Attitudes Towards Income and Wealth Inequality and Support for Scottish Independence over Time and the Interaction with National Identity." Scottish Affairs 23(1):27-54. Nisbet, Robert. 1959. "The Decline and Fall of Social Class." Pacific Sociological Review 2(1):11-28. Norman, Geoff. 2010. "Likert Scales, Levels of Measurement and the “Laws” of Statistics." Advances in Health Sciences Education 15(5):625-32. Norris, Pippa and Ronald Inglehart. 2019. Cultural Backlash: Trump, Brexit, and Authoritarian Populism. Cambridge: Cambridge University Press. Norton, Michael I. and Dan Ariely. 2011. "Building a Better America—One Wealth Quintile at a Time." Perspectives on Psychological Science 6(1):9-12. O'Connor, Kelsey J. 2017. "Who Suffered Most from the Great Recession? Happiness in the United States." RSF: The Russell Sage Foundation Journal of the Social Sciences 3(3):72-99. Oddsson, Gudmundur Aevar. 2010. "Class Awareness in Iceland." The International Journal of Sociology and Social Policy 30(5/6):292-312. OECD. 2014. "Trends in Income Inequality and Its Impact on Economic Growth." Retrieved September 25, 2017 (http://www.oecd.org/els/soc/trends-in-income-inequality-and-its-impact-on-economic-growth-SEM-WP163.pdf). OECD. 2016. "Income Inequality Remins High in the Face of Weak Recovery." Retrieved November 15, 2019 (https://www.oecd.org/social/OECD2016-Income-Inequality-Update.pdf). 147  Olssen, Mark and Michael A. Peters. 2005. "Neoliberalism, Higher Education and the Knowledge Economy: From the Free Market to Knowledge Capitalism." Journal of Education Policy 20(3):313-45. Osberg, Lars and Timothy Smeeding. 2006. ""Fair" Inequality? Attitudes toward Pay Differentials: The United States in Comparative Perspective." American Sociological Review 71(3):450-73. Pakulski, Jan and Malcolm Waters. 1996. "The Reshaping and Dissolution of Social Class in Advanced Society." Theory and Society 25(5):667-91. Pakulski, Jan and Malcolm Waters. [1996] 2008. "The Death of Class." Pp. 1022-30 in Social Stratification: Class, Race, and Gender in Sociological Perspective, edited by D. B. Grusky. Boulder: Westview Press. Payne, Keith. 2017. The Broken Ladder: How Inequality Affects the Way We Think, Live and Die. New York: Viking. Pickup, Mark and Geoffrey Evans. 2013. "Addressing the Endogeneity of Economic Evaluations in Models of Political Choice." Public Opinion Quarterly 77(3):735-54. Piketty, Thomas. 2000. "Theories of Persistent Inequality and Intergenerational Mobility." Pp. 429-76 in Handbook of Income Distribution, Vol. 1, edited by B. A. Anthony and B. François: Elsevier. Piketty, Thomas. 2014. Capital in the Twenty-First Century. Cambridge, MA: Belknap Press. Raykov, Tenko. 2011. "Intraclass Correlation Coefficients in Hierarchical Designs: Evaluation Using Latent Variable Modeling." Structural Equation Modeling: A Multidisciplinary Journal 18(1):73-90. Ridgeway, Cecilia L. 2014. "Why Status Matters for Inequality." American Sociological Review 79(1):1-16. Rodriguez-Bailon, Rosa, Boyka Bratanova, Guillermo B. Willis, Lucia Lopez-Rodriguez, Ashley Sturrock and Steve Loughnan. 2017. "Social Class and Ideologies of Inequality: How They Uphold Unequal Societies." Journal of Social Issues 73(1):99-116. Roex, Karlijn LA, Tim Huijts and Inge Sieben. 2018. "Attitudes Towards Income Inequality:‘Winners’ Versus ‘Losers’ of the Perceived Meritocracy." Acta Sociologica. Roller, Edeltraud. 1994. "Ideological Basis of the Market Economy: Attitudes toward Distribution Principles and the Role of Government in Western and Eastern Germany." European Sociological Review 10(2):105-17. Rose, David and Eric Harrison. 2007. "The European Socio-Economic Classification: A New Social Class Schema for Comparative European Research." European Societies 9(3):459-90. Rose, David and Eric Harrison, eds. 2010. Social Class in Europe : An Introduction to the European Socio-Economic Classification. Oxon: Routledge. Rözer, Jesper Jelle and Beate Volker. 2016. "Does Income Inequality Have Lasting Effects on Health and Trust?". Social Science & Medicine 149(Supplement C):37-45. Saar, Ellu. 2008. "Different Cohorts and Evaluation of Income Differences in Estonia." International Sociology 23(3):417-45. Sachweh, Patrick. 2012. "The Moral Economy of Inequality: Popular Views on Income Differentiation, Poverty and Wealth." Socio-Economic Review 10(3):419-45. Sachweh, Patrick and Sigrun Olafsdottir. 2012. "The Welfare State and Equality? Stratification Realities and Aspirations in Three Welfare Regimes." European Sociological Review 28(2):149-68. Sachweh, Patrick. 2019. "Crisis Experiences and Welfare Attitudes During the Great Recession:A Comparative Study on the Uk, Germany and Sweden." Acta Sociologica 62(2). Sands, Melissa L. 2017. "Exposure to Inequality Affects Support for Redistribution." Proceedings of the National Academy of Sciences 114(4):663-68. Schmidt-Catran, Alexander W. and Malcolm Fairbrother. 2015. "The Random Effects in Multilevel Models: Getting Them Wrong and Getting Them Right." European Sociological Review. Sealey, Anthony and Robert Andersen. 2015. "Income Inequality and Popular Support for Redistributive Policies in Canada, 1993–2008." Canadian Public Policy 41(1):51-64. 148  Shibutani, Tamotsu. 1955. "Reference Groups as Perspectives." American Journal of Sociology 60(6):562-69. Singer, Matthew. 2011. "When Do Voters Actually Think “It’s the Economy”? Evidence from the 2008 Presidential Campaign." Electoral Studies 30(4):621-32. Skeggs, Beverley and Vik Loveday. 2012. "Struggles for Value: Value Practices, Injustice, Judgment, Affect and the Idea of Class." British Journal of Sociology 63(3):472-90. Smeeding, Timothy M. 2012. "Income, Wealth, and Debt and the Great Recession." The Russell Sage Foundation and The Stanford Center on Poverty and Inequality A Great Recession Brief. Solt, Frederick. 2009. "Standardizing the World Income Inequality Database." Social Science Quarterly 90(2):231-42. Soroka, Stuart and Christopher Wlezien. 2014. "Economic Crisis and Support for Redistribution in the United Kingdom." Pp. 105-27 in Mass Politics in Tough Times: Opinions, Votes and Protest in the Great Recession. Oxford: Oxford University Press. Sosnaud, Benjamin, David Brady and Steven M. Frenk. 2013. "Class in Name Only: Subjective Class Identity, Objective Class Position, and Vote Choice in American Presidential Elections." Social Problems 60(1):81-99. Steele, Liza G. 2015. "Income Inequality, Equal Opportunity, and Attitudes About Redistribution*." Social Science Quarterly 96(2):444-64. Sullivan, Gail M. and Anthony R. Artino, Jr. 2013. "Analyzing and Interpreting Data from Likert-Type Scales." Journal of graduate medical education 5(4):541-42. Svallfors, Stefan. 1997. "Worlds of Welfare and Attitudes to Redistribution: A Comparison of Eight Western Nations." European Sociological Review 13(3):283-304. Svallfors, Stefan. 2002. "The Class Politics of Swedish Welfare Policies." Pp. 204-28 in The End of Class Politics? Class Voting in Comparative Context, edited by G. Evans. Oxford, UK: Oxford University Press. Svallfors, Stefan. 2003. "Welfare Regimes and Welfare Opinions: A Comparison of Eight Western Countries." Social Indicators Research 64(3):495-520. Svallfors, Stefan. 2004. "Class, Attitudes and the Welfare State: Sweden in Comparative Perspective." Social Policy & Administration 38(2):119-38. Svallfors, Stefan. 2005. "Class and Conformism: A Comparison of Four Western Nations." European Societies 7(2):255-86. Svallfors, Stefan. 2006. The Moral Economy of Class. Stanford: Stanford University Press. Székely, Miguel and Marianne Hilgert. 2007. "What's Behind the Inequality We Measure? An Investigation Using Latin American Data." Oxford Development Studies 35(2):197-217. Taylor-Gooby, Peter, Bjørn Hvinden, Steffen Mau, Benjamin Leruth, Mi Ah Schoyen and Adrienn Gyory. 2019. "Moral Economies of the Welfare State: A Qualitative Comparative Study." Acta Sociologica 62(2):119-34. Thompson, Jeffrey and Timothy M. Smeeding. 2011. "Inequality in the Great Recession: The Case of the United States." Political Economy Research Institute, University of Massachusetts at Amherst. Working Paper. Thompson, Jeffrey and Timothy Smeeding. 2015. "Income Inequality." Pathways (Special Issue):22-29. Torssander, Jenny and Robert Erikson. 2010. "Stratification and Mortality: A Comparison of Education, Class, Status, and Income." European Sociological Review 26(4):465-74. Trump, Kris-Stella. 2017. "Income Inequality Influences Perceptions of Legitimate Income Differences." British Journal of Political Science:1-24. UNODC. 2012. "Monitoring the Impact of Economic on Crime." Retrieved November 11, 2019 (https://www.unodc.org/documents/data-and-analysis/statistics/crime/GIVAS_Final_Report.pdf). Uslaner, Eric. 2002. The Moral Foundations of Trust Cambridge: Cambridge University Press. van der Brug, Wouter, Cees van der Eijk and Mark Franklin. 2007. The Economy and the Vote: Economic Conditions and Elections in Fifteen Countries. New York: Cambridge University Press. 149  van der Waal, Jeroen, Peter Achterberg and Dick Houtman. 2007. "Class Is Not Dead—It Has Been Buried Alive: Class Voting and Cultural Voting in Postwar Western Societies (1956–1990)." Politics & Society 35(3):403-26. Walasek, Lukasz and Gordon D. A. Brown. 2016. "Income Inequality, Income, and Internet Searches for Status Goods: A Cross-National Study of the Association between Inequality and Well-Being." Social Indicators Research 129(3):1001-14. Waters, Tony and Dagmar Waters. 2016. "Are the Terms “Socio-Economic Status” and “Class Status” a Warped Form of Reasoning for Max Weber?". Palgrave Communications 2:16002. Weatherford, M. Stephen. 1978. "Economic Conditions and Electoral Outcomes: Class Differences in the Political Response to Recession." American Journal of Political Science 22(4):917-38. Weatherford, M. Stephen. 1982. "Recessions and Social Classes: Economic Impacts and Political Opinions." Political Behavior 4(1):7-31. Weber, Max. 1922/1956/1964. "Stände Und Klassen." Pp. pp 223–27 in Wirtschaft Und Gesellschaft; Studienausgabe Erster Halbband, edited by J. W. Kiepenheuer. Tübingen, Germany: Witsch. Weber, Max. [1914] 2012. "The Distribution of Power within the Political Community: Class, Status, Party." Pp. 310-19 in Classical Sociological Theory, edited by C. Calhoun, J. Gerteis, J. Moody, S. Pfaff and I. Virk. West Sussex: Wiley-Blackwell. Weber, Max. [1922] 1978. Economy and Society. Berkeley: University of California Press. Weeden, Kim A. and David B. Grusky. 2012. "The Three Worlds of Inequality." American Journal of Sociology 117(6):1723-85. Werfhorst, Herman G. Van de and Nan Dirk de Graaf. 2004. "The Sources of Political Orientations in Post-Industrial Society: Social Class and Education Revisited." The British Journal of Sociology 55(2):211-35. Whiteley, Paul. 2016. "What Explains Electoral Responses to the ‘Great Recession’ in Europe?". Asian Journal of Comparative Politics 1(1):62-76. Wilkinson, Richard and Kate Pickett. 2010. The Spirit Level : Why Greater Equality Makes Societies Stronger. New York: Bloomsbury Press. Wilkinson, Richard and Kate Pickett. 2018. The Inner Level: How More Equal Societies Reduce Stress, Restore Sanity and Improve Everyone's Well-Being. New York: Penguin Press. Wisman, Jon D. and James F. Smith. 2011. "Legitimating Inequality: Fooling Most of the People All of the Time." American Journal of Economics and Sociology 70(4):974-1013. Wodtke, Geoffrey T. 2016. "Social Class and Income Inequality in the United States: Ownership, Authority, and Personal Income Distribution from 1980 to 2010." American Journal of Sociology 121(5):1375-415. Wodtke, Geoffrey T. 2017. "Social Relations, Technical Divisions, and Class Stratification in the United States: An Empirical Test of the Death and Decomposition of Class Hypotheses." Social Forces 95(4):1479-508. Wright, Erick Olin and Luca Perrone. 1977. "Marxist Class Categories and Income Inequality." American Sociological Review 42(1):32-55. Wright, Erik Olin. 1976. "Class Boundaries in Advanced Capitalist Societies." New Left Review 98(July-August):3-41. Wright, Erik Olin. 1993. "Class Analysis, History and Emancipation." New Left Review 202(November-December):15-35. Wright, Erik Olin. 1996. "The Continuing Relevance of Class Analysis - Comments." Theory and Society 25(5):693-716. Wright, Erik Olin. 1997. Classes. London: Verso. Wright, Erik Olin. 1998. "Rethinking, Once Again, the Concept of Class Structure." Pp. 269-348 in The Debate on Classes, edited by E. O. Wright. London: Verso. Wright, Erik Olin. 2000. Class Counts. Cambridge, UK: Cambridge University Press. Wright, Erik Olin. 2002. "The Shadow of Exploitation in Weber's Class Analysis." American Sociological Review 67(6):832-53. 150  Wright, Erik Olin. 2008. "Logics of Class Analysis." Pp. 329-49 in Social Class: How Does It Work?, edited by A. Lareau and D. Conley. New York: Russell Sage Foundation. Wu, Huiping and Shing-On Leung. 2017. "Can Likert Scales Be Treated as Interval Scales?—a Simulation Study." Journal of Social Service Research 43(4):527-32. Zhou, Xiang and Geoffrey T. Wodtke. 2018. "Income Stratification among Occupational Classes in the United States." Social Forces:945-72.  151  Appendices Appendix A Table A1. Regression results including missing categories (did not vote, not specifiable) on party support variable  (Model 1) (Model 2) (Model 3) (Model 4) (Model 5) (Model 6) Lower managers/professionals, higher supervisory/technicians 0.146***  0.121*** 0.136*** .112*** # 0.085* #  (0.02)  (0.02) (0.02) (.02) (0.04)        Intermediate occupations 0.220***  0.169*** 0.154*** .158*** # 0.167*** #  (0.02)  (0.02) (0.04) (.02) (0.04)        Small employers and self-employed 0.196***  0.135*** 0.217** .132*** 0.147**  (0.02)  (0.02) (0.08) (.02) (0.06)        Lower supervisors and technicians 0.256***  0.199*** 0.145** .189*** # 0.236***  (0.02)  (0.02) (0.05) (.02) (0.05)        Lower sales and service 0.287***  0.221*** 0.224*** .201*** # 0.196***  (0.02)  (0.02) (0.05) (.02) (0.05)        Lower technical 0.320***  0.253*** 0.129 .233*** # 0.192***  (0.02)  (0.02) (0.09) (.02) (0.05)        Routine 0.315***  0.245*** 0.211** .222*** # 0.207*** #  (0.02)  (0.02) (0.08) (.02)  (0.04)        152  Female 0.103*** 0.092*** 0.105*** 0.106*** .098*** # 0.095*** #  (0.01) (0.01) (0.01) (0.01) (.01) (0.01)        Age 0.003*** 0.003*** 0.003*** 0.003*** .003*** # 0.003*** #  (0.00) (0.00) (0.00) (0.00) (.00) (0.00)        Married -0.002 0.005 -0.001 -0.001 .002 0.003  (0.01) (0.01) (0.01) (0.01) (.01) (0.01)        Not religious 0.073*** 0.069*** 0.077*** 0.077*** .048*** 0.046***  (0.01) (0.01) (0.01) (0.01) (.01) (0.01)        Protestant 0.004 -0.002 0.001 0.002 .007 0.009  (0.01) (0.01) (0.01) (0.01) (.01) (0.01)        Other religion 0.039* 0.045** 0.042* 0.042* .025 0.024  (0.02) (0.02) (0.02) (0.02) (.02) (0.02)        1999 0.107*** 0.109*** 0.114*** 0.114*** .122*** # 0.122*** #  (0.01) (0.01) (0.01) (0.01) (.01) (0.01)        2009 0.080*** 0.092*** 0.094*** 0.093*** .096*** # 0.095*** #  (0.01) (0.01) (0.01) (0.01) (.01) (0.01)         (0.03) (0.02) (0.03) (0.03)  (0.03)        Less than secondary  0.230*** 0.127*** 0.142 .128*** # 0.134*** #   (0.01) (0.02) (0.07) (.02) (0.02) 153         Secondary/Higher Secondary  0.205*** 0.127*** 0.124*** .132*** 0.136***   (0.01) (0.01) (0.03) (.01) (0.01)        Above higher secondary  0.123*** 0.081*** 0.101** .090*** 0.094***   (0.01) (0.02) (0.04) (.02) (0.02)        Lower managers/professionals, higher supervisory/technicians*Less than secondary    0.039       (0.08)          Lower managers/professionals, higher supervisory/technicians*Secondary/Higher Secondary    -0.039       (0.04)          Lower managers/professionals, higher supervisory/technicians*Above higher secondary    -0.011       (0.04)          Intermediate occupations*Less than secondary    0.049       (0.09)          Intermediate occupations*Secondary/Higher Secondary    0.030       (0.05)          Intermediate occupations*Above higher secondary    -0.039       (0.06)          Small employers and self-employed*Less than secondary    -0.126       (0.11)          Small employers and self-employed*Secondary/Higher Secondary    -0.066       (0.09)   154         Small employers and self-employed*Above higher secondary    -0.100       (0.10)          Lower supervisors and technicians*Less than secondary    0.075       (0.09)          Lower supervisors and technicians*Secondary/Higher Secondary    0.059       (0.06)          Lower supervisors and technicians*Above higher secondary    0.026       (0.07)          Lower sales and service*Less than secondary    0.033       (0.09)          Lower sales and service*Secondary/Higher Secondary    -0.019       (0.06)          Lower sales and service*Above higher secondary    -0.001       (0.07)          Lower technical*Less than secondary    0.084       (0.12)          Lower technical*Secondary/Higher Secondary    0.149       (0.10)          Lower technical*Above higher secondary    0.089       (0.11)          Routine*Less than secondary    0.008       (0.11)   155         Routine*Secondary/Higher Secondary    0.057       (0.09)          Routine*Above higher secondary    -0.029       (0.10)          Left     .154*** 0.233***      (.01) (0.04)        Right     -.179*** -0.389***      (.02) (0.04)        Other, or no party specified     .047** 0.165***      (.02) (0.05)        Did not vote     .050*** 0.095*      (.01) (0.04)        Lower managers/professionals, higher supervisory/technicians*Left      -0.012       (0.05)        Lower managers/professionals, higher supervisory/technicians*Right      0.168**       (0.05)        Lower managers/professionals, higher supervisory/technicians*other, or no party specified      -0.061       (0.06)        Lower managers/professionals, higher supervisory/technicians*Did not vote      -0.022       (0.05)        156  Intermediate occupations*Left      -0.083       (0.05)        Intermediate occupations*Right      0.187**       (0.06)        Intermediate occupations*other, or no party specified      -0.196**       (0.06)        Intermediate occupations*Did not vote      -0.040       (0.05)        Small employers and self-employed*Left      -0.161*       (0.07)        Small employers and self-employed*Right      0.211**       (0.08)        Small employers and self-employed*other, or no party specified      -0.153*       (0.08)        Small employers and self-employed*Did not vote      -0.061       (0.07)        Lower supervisors and technicians*Left      -0.140*       (0.06)        Lower supervisors and technicians*Right      0.162**       (0.06)        Lower supervisors and technicians*other, or no party specified      -0.203**       (0.07)        157  Lower supervisors and technicians*Did not vote      -0.112*       (0.06)        Lower sales and service*Left      -0.112*       (0.05)        Lower sales and service*Right      0.249***       (0.06)        Lower sales and service*other, or no party specified      -0.168*       (0.07)        Lower sales and service*Did not vote      -0.022       (0.05)        Lower technical*Left      -0.081       (0.05)        Lower technical*Right      0.388***       (0.06)        Lower technical*other, or no party specified      -0.101       (0.06)        Lower technical*Did not vote      -0.028       (0.05)        Routine*Left      -0.126*       (0.05)        Routine*Right      0.379***       (0.06)        158  Routine*other, or no party specified      -0.101       (0.06)        Routine*Did not vote      -0.062       (0.05)        Constant 3.250*** 3.320*** 3.206*** 3.203*** 3.216 3.226***  (0.03) (0.02) (0.03) (0.03) (.03) (0.04) Number of Individuals 63595 79216 63105 63105 63105 63105 Number of Countries 27 27 27 27 27 27 Adjusted R2 0.157 0.152 0.159 0.160 .171 0.174 Note: The “#” symbol indicates which coefficients are significantly different from model 5 and 6 in Table 2.2.    159  Table A.2. Odds ratios from ordered logistic regression models and interactions between class, education, and political affiliation in inequality views Class (Model 1) (Model 2) (Model 3) (Model 4) (Model 5) (Model 6) Lower managers/professionals, higher supervisory/technicians 1.398***  1.332*** 1.460*** 1.303*** 1.126  (0.06)  (0.06) (0.09) (0.05) (0.09)        Intermediate occupations 1.605***  1.455*** 1.569*** 1.425*** 1.297**  (0.07)  (0.07) (0.15) (0.07) (0.13)        Small employers and self-employed 1.572***  1.360*** 1.530 1.404*** 1.260  (0.10)  (0.09) (0.36) (0.09) (0.17)        Lower supervisors and technicians 1.844***  1.618*** 1.408** 1.579*** 1.535***  (0.09)  (0.08) (0.18) (0.08) (0.15)        Lower sales and service 1.890***  1.638*** 1.658*** 1.550*** 1.406***  (0.09)  (0.08) (0.23) (0.08) (0.14)        Lower technical 2.200***  1.873*** 1.215 1.781*** 1.423***  (0.11)  (0.10) (0.35) (0.09) (0.14)        Routine 2.264***  1.913*** 1.582* 1.769*** 1.509***  (0.10)  (0.09) (0.32) (0.09) (0.14)        female 1.291*** 1.276*** 1.289*** 1.288*** 1.272*** 1.265***  (0.03) (0.03) (0.03) (0.03) (0.03) (0.03)        Age 1.010*** 1.007*** 1.008*** 1.008*** 1.008*** 1.008***  (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)        Married 0.995 0.980 0.998 0.996 1.009 1.010  (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) 160         Not religious 1.296*** 1.293*** 1.314*** 1.317*** 1.200*** 1.188***  (0.05) (0.05) (0.05) (0.05) (0.04) (0.04)        Protestant 0.988 0.985 0.989 0.991 1.001 1.003  (0.04) (0.04) (0.04) (0.04) (0.04) (0.04)        Other religion 1.174** 1.189** 1.186** 1.186** 1.107 1.106  (0.07) (0.07) (0.07) (0.07) (0.07) (0.07)        1999 1.195*** 1.231*** 1.221*** 1.222*** 1.217*** 1.217***  (0.04) (0.04) (0.04) (0.04) (0.04) (0.04)        2009 1.104** 1.154*** 1.143*** 1.143*** 1.133*** 1.133***  (0.04) (0.04) (0.04) (0.04) (0.04) (0.04)        Education       Less than secondary  1.984*** 1.436*** 1.334 1.466*** 1.491***   (0.08) (0.07) (0.27) (0.07) (0.07)        Secondary/Higher Secondary  1.597*** 1.280*** 1.375*** 1.306*** 1.322***   (0.05) (0.05) (0.10) (0.05) (0.05)        Above higher secondary  1.299*** 1.150** 1.275* 1.191*** 1.206***   (0.06) (0.05) (0.12) (0.05) (0.05) Class*Education Interactions       Lower managers/professionals, higher supervisory/technicians*Less than secondary    1.241       (0.28)          Lower managers/professionals, higher supervisory/technicians*Secondary/Higher Secondary    0.795*       (0.07)   161         Lower managers/professionals, higher supervisory/technicians*Above higher secondary    0.904       (0.11)          Intermediate occupations*Less than secondary    1.020       (0.25)          Intermediate occupations*Secondary/Higher Secondary    0.898       (0.11)          Intermediate occupations*Above higher secondary    0.820       (0.12)          Small employers and self-employed*Less than secondary    0.814       (0.26)          Small employers and self-employed*Secondary/Higher Secondary    0.943       (0.24)          Small employers and self-employed*Above higher secondary    0.818       (0.25)          Lower supervisors and technicians*Less than secondary    1.297       (0.32)          Lower supervisors and technicians*Secondary/Higher Secondary    1.148   162      (0.17)          Lower supervisors and technicians*Above higher secondary    0.981       (0.19)          Lower sales and service*Less than secondary    1.293       (0.33)          Lower sales and service*Secondary/Higher Secondary    0.909       (0.14)          Lower sales and service*Above higher secondary    0.890       (0.18)          Lower technical*Less than secondary    1.649       (0.58)          Lower technical*Secondary/Higher Secondary    1.543       (0.46)          Lower technical*Above higher secondary    1.330       (0.46)          Routine*Less than secondary    1.314       (0.38)          Routine*Secondary/Higher Secondary    1.206       (0.26)          Routine*Above higher secondary    1.061       (0.26)   163  Party Affiliation       Left     1.473*** 1.605***      (0.05) (0.14)        Right     0.729*** 0.504***      (0.02) (0.04) Class*Party Interactions       Lower managers/professionals, higher supervisory/technicians*Left      1.055       (0.11)        Lower managers/professionals, higher supervisory/technicians*Right      1.351**       (0.14)        Intermediate occupations*Left      0.917       (0.11)        Intermediate occupations*Right      1.356*       (0.16)        Small employers and self-employed*Left      0.866       (0.15)        Small employers and self-employed*Right      1.461*       (0.23)        Lower supervisors and technicians*Left      0.803       (0.10)        Lower supervisors and technicians*Right      1.303*       (0.16)        164  Lower sales and service*Left      0.843       (0.10)        Lower sales and service*Right      1.523***       (0.19)        Lower technical*Left      0.925       (0.11)        Lower technical*Right      2.002***       (0.25)        Routine*Left      0.835       (0.09)        Routine*Right      1.976***       (0.24) Cut 1 -2.420 -2.611 -2.363 -2.335 -2.530 -1.845  (.08) (.08) (.08) (.08) (.10) (.09) Cut 2 -.525 -.719 -.466 -.437 -.610 .075  (.07) (.07) (.07) (.07) (.09) (.08) Cut 3 .429 .229 .488 .518 .367 1.052  (.07) (.06) (.07) (.07) (.09) (.08) Cut4 2.509 2.300 2.572 2.604 2.495 3.180  (.07) (.07) (.07) (.08) (.09) (.08) N 33678 33678 33678 33678 33678 33678 Number of countries 26 26 26 26 26 26 Log lik. -37730.829 -37822.358 -37693.296 -37669.873 -37307.924 -37242.310 Chi-squared 4700.349 4626.903 4735.668 4802.860 5203.689 5398.845 Exponentiated coefficients; Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001  165  Table A.3 Joint tests of significance for Models 1 through 6 in Chapter 1 Model 1 Model 2 Model 3 Model 4 Model 5 Model 6       Class*** Education*** Class*** Class*** Class*** Class*** Female*** Female*** Education*** Education*** Education*** Education*** Age*** Age*** Female*** Class X Education** Party Support*** Party Support*** Religion*** Religion*** Age*** Female Female Class X Party support*** Married Married Religion*** Age*** Age*** Female Year*** Year*** Married Religion*** Religion*** Age***   Year*** Married Married Religion**    Year*** Year*** Married      Year*** ** p < 0.01, *** p < 0.001       166  Analysis of missing data Political orientation is derived by data disseminators from two questions: 1) whether a respondent voted in their respective country’s last national election; and 2) the country-specific party a respondent voted for in last their respective country’s last national election. Using the second question, data disseminators then place individuals on an ideological left-right continuum. This means that respondents must have first voted in their country’s last national election and second, answered a question on which particular political party they voted for in order to be included in the party support variable. Thus, those who did not vote, were not eligible to vote, or did not divulge which political party they voted for are considered missing. Exploring how class and education relate to inequality views among the politically inactive would be a useful direction for future research. Table A4 below provides information on the percentage of non-responses of inequality views, party support and education of the various classes. There are low numbers of missing values on the inequality views of the various classes. For instance, for the routine classes 2.9 per cent of inequality views are missing. Missing values for education are also low across the class spectrum, with “lower sales and service” classes having the highest percentage of missing at 1.20. In contrast, the various classes combined missing values on the party support variable range from 53.56 per cent in the routine classes, to 39.86 per cent in the Large employers, higher managers/professionals. There are high numbers of missing values for party support across the classes, yet these are highest for the routine classes which may mean that the political orientations of this class could be underrepresented.  Table A5 provides the percentages of missing values for inequality views and party support across levels of education. Those with “less than secondary” education levels have higher percentages of missing values for both inequality views (4.58) and party support (54.18). There are high numbers of missing values for party support across various education levels, yet they are highest for those with “less than secondary” levels of education, which could mean that the political orientations of this group may also be underrepresented.  167  Table A7, provides the percentage of missing cases for party support across countries in each time period. Missing values for party support are lowest in the United States in 1999 at 2.12 per cent and highest in Russia in 2009 at 76.1 percent. Further, data for the party support variable are missing in some countries and in some years. For instance, in 1992 values for the party support variable were missing from Canada, Chile, Cyprus, France, Hungary, Israel, Italy, Japan, Latvia, New Zealand, Philippines, Poland, Portugal, Spain, Sweden, and Switzerland. In 1999, party affiliation was missing from the following countries: Israel, Italy, Latvia, Sweden and Switzerland. In 2009, values were missing from Canada, Cyprus, Hungary, Israel, New Zealand and Poland. There are, of course, differences in voter turnout across countries which may account for the high numbers of missing values. For instance, in the most recent Canadian federal election only 65.95 per cent of eligible voters cast a ballot (Global News 2019). Moreover, an unwillingness to divulge particular party support in countries where doing so could have personal ramifications (such as in Russia) may also account for high numbers of missing values.  The above analyses demonstrate that the missing data appear to be highly concentrated on the party support variable. Due to the high number of missing values on party support, including the variable in model 6 (which includes an interaction between class and vote) causes the sample size to decline from the previous models. Tests of statistical significance could be misleading if changes in significance levels are the result of a decline in the sample size and not because of an effect of the party support variable. To maintain a consistent sample size across the models, I constructed a variable “touse”, coded 1 if there is no missing data on any of the variables, otherwise 0. I then ran the regression models limited to this variable which includes cases with the nonzero variables on “touse”. In other words, the models are limited to cases with data on all variables. 168  A.4 Percentage of non-responses for inequality views, party support and education by class   Party Support    Class Inequality Views (Other) (Not applicable) (Total Missing) Education   Large Employer, higher managers/professionals 1.89 10.76  29.10  39.86 .72    Lower managers/professional, higher supervisory/technicians  1.76 11.05  30.43  41.48 .78    Intermediate 1.99 13.63  30.35  43.48 .89    Small employers and self-employed 2.06 15.77  41.28  57.05 .86    Lower supervisor and technician 2.23 13.11  32.49  45.60 .89    Lower sales and service 2.46 15.84  35.03  50.87 1.20    Lower technical 2.48 16.75  35.81  52.56 .65    Routine 2.90 17.45  36.11  53.56 .82    Note: Survey 1987 dropped from analysis            A.5 Percentage of non-responses for inequality views and party support by education   Party Support Education Inequality Views (other) (not applicable) (Total Missing) Less than Secondary  4.58 17.32  36.86 54.18 Secondary/higher secondary 2.27 15.18  34.52 49.70 Above higher secondary 2.16 15.20  32.80 48 University/graduate studies 1.86 12.77  29.61 42.38 Note: Survey 1987 dropped from analysis          169  Table A6. Percentage of non-responses for inequality views and party support by class Class Left Centre Right Other, no specification Did not vote   Large Employer, higher managers/professionals 22.54 15.17  22.43  10.76 29.1    Lower managers/professional, higher supervisory/technicians  24.4 14.61  19.51  11.05 30.43    Intermediate 23.89 13.32  18.8  13.63 30.35    Small employers and self-employed 13.44 12.1  17.42  15.77 41.28    Lower supervisor and technician 22.04 12.9  19.46  13.11 32.49    Lower sales and service 22.32 11.96  14.85  15.84 35.03    Lower technical 21.33 12.12  13.99  16.75 35.81    Routine 22.58 12.12  11.75  17.45 36.11    Note: Survey 1987 dropped from analysis               170  Table A.7 Percentage of missing cases for party support by country and year Country Survey Year N Total Missing Australia 1992 2203 10.21  1999 1672 21.29  2009 1525 22.16 Austria 1992 1027 28.92  1999 1016 40.06  2009 1019 40.72 Bulgaria 1992 1198 72.46 1999 1102 58.53  2009 1000 61.70 Canada 1999 974 36.65 Chile 1999 1503 50.69 Cyprus 1999 1000 22.60 Czech Republic 1992 678 3.24 1999 1834 38.01 2009 1205 44.98 France 1999 1889 17.42  2009 2817 24.49 Germany 1992 3391 36.13  1999 1432 39.17  2009 1395 11.40 Hungary 1999 1208 45.7 Israel - - - Italy 2009 1084 50 Japan 1999 1325 63.62  2009 1296 46.60 Latvia 2009 1069 70.25 New Zealand 1999 1108 32.92 Norway 1992 1538 31.86  1999 1268 15.46  2009 1456 17.92 Philippines 1999 1200 59.84  2009 1200 56.50 Poland 1999 1135 42.73 Portugal 1999 1144 34.52  2009 1000 50.40 Russia 1992 1983 56.53  1999 1705 54.48  2009 1603 76.10 Slovak Republic 1992 422 2.13 1999 1082 50.74 2009 1159 57.72 171  Slovenia 1992 1049 56.34  1999 1006 53.38  2009 1065 44.51 Spain 1999 1211 23.70  2009 1215 22.06 Sweden 2009 1137 13.64     Switzerland 2009 1229 42.31 UK 1992 1066 11.44  1999 804 20.02  2009 958 25.05 US 1992 1273 3.14  1999 1272 2.12 2009 1581 2.53 Note: Columns labelled “N” and “Missing” include all types of missing values.172  Appendix B Table B.1 Odds ratios from ordered logistic regression models and interactions between year, class and political affiliation in inequality views  (Model 1) (Model 2) (Model 3) (Model 4) (Model 5)       1992 0.937* 0.885*** 0.776*** 0.809*** 0.731***  (0.03) (0.03) (0.04) (0.04) (0.06)       1999 1.101** 1.076* 1.050 1.134* 1.207*  (0.03) (0.03) (0.05) (0.06) (0.09)       Female 1.267*** 1.261*** 1.261*** 1.261*** 1.253***  (0.03) (0.03) (0.03) (0.03) (0.03)       Age 1.010*** 1.008*** 1.008*** 1.008*** 1.008***  (0.00) (0.00) (0.00) (0.00) (0.00)       Married 0.961 1.009 1.007 1.009 1.009  (0.02) (0.02) (0.02) (0.02) (0.02)       Not religious 1.233*** 1.194*** 1.195*** 1.193*** 1.185***  (0.04) (0.04) (0.04) (0.04) (0.04)       Protestant 0.978 1.001 1.000 1.002 1.003  (0.04) (0.04) (0.04) (0.04) (0.04)       Other religion 1.160* 1.111 1.112 1.110 1.109  (0.07) (0.07) (0.07) (0.07) (0.07)       Class      Working Class  1.404*** 1.297*** 1.404*** 1.985***   (0.04) (0.06) (0.04) (0.18)       Intermediate Class  1.228*** 1.158** 1.230*** 1.458***   (0.04) (0.06) (0.04) (0.12) Education      Less than secondary  1.548*** 1.560*** 1.547*** 1.585***   (0.07) (0.07) (0.07) (0.08)       Secondary/Higher Secondary  1.362*** 1.381*** 1.362*** 1.396***   (0.05) (0.05) (0.05) (0.05)       Above higher secondary  1.232*** 1.245*** 1.230*** 1.255***   (0.05) (0.05) (0.05) (0.05) 173  Party Affiliation      Left  2.031*** 2.032*** 2.045*** 2.920***   (0.06) (0.06) (0.09) (0.21)       Centre  1.376*** 1.375*** 1.336*** 1.687***   (0.04) (0.04) (0.07) (0.15) Two-way interactions      Working Class*1992   1.281***  1.249    (0.09)  (0.15)       Working Class*1999   1.027  0.865    (0.07)  (0.10)       Intermediate Class*1992   1.148  1.037    (0.08)  (0.12)       Intermediate Class*1999   1.059  0.883    (0.07)  (0.10)       Left*1992    1.111 1.037     (0.07) (0.11)       Left*1999    0.909 0.810*     (0.06) (0.08)       Centre*1992    1.199* 1.195     (0.09) (0.15)       Centre*1999    0.948 0.819     (0.07) (0.10)       Working Class*Left     0.477***      (0.05)       Working Class*Centre     0.657**      (0.09)       Intermediate Class*Left     0.690***      (0.07)       Intermediate Class*Centre     0.707**      (0.09) Three-way interactions      Working Class*Left*1992     1.114      (0.18)       174  Working Class*Left*1999     1.364*      (0.20)       Working Class*Centre*1992     0.876      (0.16)       Working Class*Centre*1999     1.073      (0.19)       Intermediate Class*Left*1992     1.152      (0.19)       Intermediate Class*Left*1999     1.121      (0.17)       Intermediate Class*Centre*1992     1.223      (0.24)       Intermediate Class*Centre*1999     1.665**      (0.30)       Cut 1 -2.966 -2.351 -2.394 -2.357 -2.211  (.08) (.082) (.08) (.09) (.09) Cut2 -1.082 -.441 -.483 -.445 -.292  (.07) (.07) (.07) (.08) (.08) Cut 3 -.141 .528 .486 .524 .683  (.07) (.07) (.07) (.08) (.08) Cut 4 1.914 2.646 2.606 2.643 2.810  (.07) (.07) (.08) (.08) (.08) Number of Individuals 33678 33678 33678 33678 33678 Number of countries 26 26 26 26 26 Log lik. -38008.605 -37341.071 -37332.101 -37333.852 -37258.596 Chi-squared 4463.389 5131.945 5186.603 5167.549 5390.142 Exponentiated coefficients; Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001      175     Table B.2 Regression results including missing categories (did not vote, not specifiable) on party support variable  (Model 1) (Model 2) (Model 3) (Model 4) (Model 5) Survey Year      1992 -0.068***  -0.095*** #  -0.163***  -0.141*** # -0.234*** #  (0.01) (0.01) (0.02) (0.03) (0.04)       1999 0.029***  0.027**  -0.000 0.070** 0.086*  (0.01) (0.01) (0.02) (0.02) (0.04) Controls      female 0.092*** 0.097***  0.098*** # 0.097*** # 0.095*** #  (0.01) (0.01) (0.01) (0.01) (0.01)       Age 0.003*** 0.003***  0.003*** # 0.003*** # 0.003*** #  (0.00) (0.00) (0.00) (0.00) (0.00)       married -0.003  0.003  0.003 0.003 0.004  (0.01) (0.01) (0.01) (0.01) (0.01)       Not religious 0.056***  0.047***  0.047*** 0.045*** 0.044***  (0.01) (0.01) (0.01) (0.01) (0.01)       Protestant -0.000  0.008  0.008 0.008 0.009  (0.01) (0.01) (0.01) (0.01) (0.01)       Other religion 0.038*  0.025  0.025 0.027 0.026  (0.02) (0.02) (0.02) (0.02) (0.02) Class      Working Class  0.139*** # 0.093***  0.139*** # 0.344***   (0.01) (0.02) (0.01) (0.04)       Intermediate Class  0.084*** # 0.057***  0.085*** # 0.202***   (0.01) (0.02) (0.01) (0.04) Education      Less than secondary  0.145*** # 0.147*** # 0.145*** # 0.153*** #   (0.02) (0.02) (0.02) (0.02) 176        Secondary/Higher Secondary  0.149***  0.155*** 0.149*** 0.158***   (0.01) (0.01) (0.01) (0.01)       Above higher secondary  0.105*** 0.109*** 0.104*** 0.110***   (0.02) (0.02) (0.02) (0.02) Party Affiliation      Left  0.336***  0.336***  0.334*** 0.520***   (0.01) (0.01) (0.02) (0.03)       Centre  0.181***  0.181***   0.170*** 0.292***   (0.02) (0.02) (0.02) (0.04)       Other, or no party specified  0.226*** 0.228*** 0.195*** 0.362***   (0.01) (0.01) (0.02) (0.04)       Did not vote  0.231*** 0.231*** 0.267*** 0.409***   (0.01) (0.01) (0.02) (0.03) Two-way interactions      Working Class*1992   0.109*** #  0.192**    (0.02)  (0.06)       Working Class*1999   0.050*  -0.045    (0.02)  (0.06)       Intermediate Class*1992   0.080**  0.072    (0.03)  (0.07)       Intermediate Class*1999   0.020  -0.039    (0.02)  (0.06)       Left*1992    0.099** 0.095     (0.03) (0.06)       Left*1999    -0.060* # -0.094 #     (0.03) (0.05)       Centre*1992    0.099** 0.125     (0.04) (0.07)       Centre*1999    -0.045 # -0.111 #     (0.03) (0.06)       Other, or no party specified*1992    0.104* 0.178**     (0.04) (0.07) 177        Other, or no party specified*1999    0.028 0.022     (0.03) (0.06)       Did not vote*1992    -0.009 0.045     (0.03) (0.06)       Did not vote*1999    -0.093** -0.160**     (0.03) (0.05)       Working Class*Left     -0.367***      (0.05)       Working Class*Centre     -0.238***      (0.06)       Working Class*Other, or no party specified     -0.289***      (0.05)       Working Class*Did not vote     -0.285***      (0.05)       Intermediate Class*Left     -0.213***      (0.05)       Intermediate Class*Centre     -0.159**      (0.06)       Intermediate Class*Other, or no party specified     -0.227***      (0.06)       Intermediate Class*Did not vote     -0.149**      (0.05) Three-way interactions      Working Class*Left*1992     -0.028      (0.08)       Working Class*Left*1999     0.098      (0.07)       Working Class*Centre*1992     -0.099      (0.09)       Working Class*Centre*1999     0.049 178       (0.08)       Working Class*Other, or no party specified*1992     -0.179      (0.09)       Working Class*Other, or no party specified*1999     0.042      (0.08)       Working Class*Did not vote*1992     -0.131      (0.07)       Working Class*Did not vote*1999     0.159*      (0.07)       Intermediate Class*Left*1992     0.049      (0.08)       Intermediate Class*Left*1999     0.049      (0.07)       Intermediate Class*Centre*1992     0.042      (0.10)       Intermediate Class*Centre*1999     0.204*      (0.08)       Intermediate Class*Other, or no party specified*1992     -0.022      (0.10)       Intermediate Class*Other, or no party specified*1999     0.020      (0.08)       Intermediate Class*Did not vote*1992     -0.031      (0.08)       Intermediate Class*Did not vote*1999     0.029      (0.07)       Intercept 3.534*** 3.196*** 3.221*** 3.191*** 3.092***  (0.02) (0.03) (0.03) (0.03) (0.04) 179  Number of Individuals 79844 63105 63105 63105 63105 Number of Countries 27 27 27 27 27 Adjusted R2 0.145 0.170 0.170 0.171 0.174 Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Note: The “#” symbols next to the coefficients indicate which estimates significantly differ from models 2 – 5 in Table. 3.4       Table B.3 Joint tests of significance for Models 1 through 5 Model 1 Model 2 Model 3 Model 4 Model 5 Year*** Class*** Class*** Class*** Class*** Female*** Education*** Class X Year*** Education*** Class X Year** Age*** Party Support*** Education*** Party Support*** Class X Party Support*** Religion*** Year*** Party Support*** Party Support X Year*** Education** Married* Female*** Year*** Year*** Party Support***  Age*** Female*** Female*** Party Support X Year**  Religion*** Age*** Age*** Class X Party Support X Year  Married Religion*** Religion*** Year***   Married Married Female***     Age***     Religion***     Married P* < .05, ** p < 0.01, *** p < 0.00   180  Appendix C Model 1 in table C.1 determines the proportions of variability in inequality views attributable to the individual and country-year levels. Model 2 includes only the country-level variables. Model 3 removes the country-level factors and replaces them with individual-level characteristics. Model 4 includes both the country-level and individual-level variables together.  To begin I calculated the intra-class correlation for model 1 which reveals the ratio of the between-cluster variance to the total variance. This is useful as it enables me to understand how much of the overall variation in inequality views is explained by clustering at the country-year level (Raykov 2011). Inequality views are slightly correlated with the country-year effects; approximately 13 per cent of the overall variation in responses is explained by clustering at the country-year level. In model 1, we also see that average inequality views are 4.196. The random intercept variance for inequality views is significant .12 (p < .001), indicating that inequality views vary significantly across country-years.  Model 2 adds the country-level factors. Adding the country-level components to the model reduces the intra-class correlation to .06, which means that conditional on the country-level components, approximately 6 per cent of the overall variation in the response is explained by clustering at the country-year level. The reduction in the random variance indicates that the inclusion of these societal factors may be important in moderating the relationship between class and inequality views.  Model 3 removes the country-level factors and replaces them with individual-level characteristics. Conditional on individual-level factors, approximately 11 per cent of the overall variation in the response is explained by clustering at the country-year level. The country-year variance is lower than it was in the null-model, which suggests that the various individual-level factors may also be important in shaping inequality views.  Model 4 includes both the country-level and individual-level variables together. When both induvial and societal-level factors are included, approximately 5 per cent of the overall variation in the 181  response is explained by clustering at the country-year level. The random component for the country-years remain statistically significant. Further, the lower Akaike information criterion (AIC) values compared to the previous models, further justify the inclusion of both country and individual levels factors.  182  Table C.1. Models assessing variability in inequality views attributable to the individual and country-year levels  (Model 1) (Model 2) (Model 3) (Model 4) Societal-level     Gini  0.872  0.953   (0.77)  (0.71)      GDP  -0.002  -0.011   (0.03)  (0.02)      Conservative (ref: Liberal)  0.465***  0.398***   (0.10)  (0.09)      Social Democratic  -0.124  -0.120   (0.14)  (0.13)      Former Communist  0.598***  0.510***   (0.11)  (0.10)      Mediterranean  0.375**  0.237*   (0.13)  (0.12)      1992  -0.159 -0.140 -0.152   (0.08) (0.11) (0.08)      1999  -0.108 -0.070 -0.083   (0.07) (0.10) (0.06) Individual-level      Working Class (ref: Salariat)   0.156*** 0.155***    (0.01) (0.01)      Intermediate Class   0.107*** 0.107***    (0.01) (0.01)      Less than secondary    0.206*** 0.206*** (ref: University/postgraduate)   (0.02) (0.02)      Secondary/Higher Secondary   0.172*** 0.171***    (0.02) (0.02)      Above higher secondary   0.104*** 0.105***    (0.02) (0.02)      Left (ref: right)   0.351*** 0.350***    (0.01) (0.01)      183  Centre   0.200*** 0.199***    (0.01) (0.01)      Female   0.109*** 0.109***    (0.01) (0.01)      Age   0.003*** 0.003***    (0.00) (0.00)      Married   0.001 0.001    (0.01) (0.01)      Not religious (ref: Catholic)   0.010 0.012    (0.01) (0.01)      Protestant   -0.024 -0.019    (0.02) (0.02)      Other religion   0.030 0.027    (0.02) (0.02)      Intercept 4.196*** 4.020*** 3.627*** 3.677***  (0.05) (0.77) (0.07) (0.71) Random Effects     Country-year .12*** .05*** .09*** .04***  (0.02) (0.01) (0.02) (0.01) Number of individuals 32627 32627 32627 32627 Number of countries 24 24 24 24 Number of country-years 53 53 53 53 AIC 83930 83898 82154 82122 Intraclass correlation .13 (.02) .06 (.01)  .11 (.02) .05 (.01) Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 184  Table C.2 Joint tests of significance for Models 1 through 5 Model 1 Model 2 Model 3 Model 4 Model 5 Class*** Class*** Class*** Class* Class* Education*** Education*** Education*** Education*** Education*** Party *** Party*** Party*** Party*** Party*** Female*** Female*** Female*** Female*** Female*** Age*** Age*** Age*** Age*** Age*** Married Married Married Married Married Religion Religion Religion Religion Religion Gini Gini* Gini Gini Gini GDP GDP GDP GDP GDP Regime Type*** Regime Type*** Regime Type*** Regime Type*** Regime Type*** Year Year Year Year Year  Class X Gini*** Class X GDP Class X Regime Type*** Class X Gini***     Class X Regime Type*** * p < .05, ** p < 0.01, *** p < 0.001   185  Table C.3. Regression results including missing categories (did not vote, not specifiable) on party support variable  (Model 1) (Model 2) (Model 3) (Model 4) (Model 5) Individual-level      Working Class (ref: Salariat) 0.138*** 0.107*** -0.184 0.140*** 0.170***  (0.02) (0.02) (0.39) (0.04) (0.04)       Intermediate Class 0.071*** 0.053*** -0.187 0.060* 0.077**  (0.01) (0.01) (0.29) (0.03) (0.03)       Less than secondary 0.205*** 0.205*** 0.205*** 0.207*** 0.207*** (ref: University/Postgraduate) (0.03) (0.03) (0.03) (0.03) (0.03)       Secondary/Higher Secondary 0.200*** 0.201*** 0.200*** 0.200*** 0.201***  (0.03) (0.03) (0.03) (0.03) (0.03)       Above higher secondary 0.116*** 0.117*** 0.116*** 0.116*** 0.117***  (0.02) (0.02) (0.02) (0.02) (0.02)       Left (ref: right) 0.346*** 0.346*** 0.346*** 0.345*** 0.345***  (0.04) (0.04) (0.04) (0.03) (0.03)       Centre 0.198*** 0.199*** 0.198*** 0.197*** 0.197***  (0.04) (0.04) (0.04) (0.04) (0.04)       Other, or no party specified 0.240*** 0.240*** 0.240*** 0.240*** 0.239***  (0.03) (0.03) (0.03) (0.03) (0.03)       Did not vote 0.235*** 0.235*** 0.235*** 0.235*** 0.234***  (0.03) (0.03) (0.03) (0.03) (0.03)       Female 0.092*** 0.093*** 0.092*** 0.092*** 0.092***  (0.01) (0.01) (0.01) (0.01) (0.01)       186  Age 0.003*** 0.003*** 0.003*** 0.003*** 0.003***  (0.00) (0.00) (0.00) (0.00) (0.00)       Married 0.010 0.010 0.010 0.010 0.010  (0.01) (0.01) (0.01) (0.01) (0.01)       Not religious (ref: Catholic) 0.005 0.005 0.005 0.004 0.005  (0.02) (0.02) (0.02) (0.02) (0.02)       Protestant -0.016 -0.016 -0.016 -0.017 -0.016  (0.01) (0.01) (0.01) (0.01) (0.01)       Other religion 0.007 0.007 0.007 0.007 0.007  (0.03) (0.03) (0.03) (0.03) (0.03) Societal-level      Gini 0.822 1.815* 0.821 0.799 1.574*  (0.64) (0.75) (0.64) (0.63) (0.74)       GDP 0.020 0.021 0.013 0.021 0.021  (0.02) (0.02) (0.02) (0.02) (0.02)       Conservative (ref: Liberal) 0.439*** 0.440*** 0.439*** 0.375*** 0.425***  (0.08) (0.08) (0.08) (0.09) (0.10)       Social Democratic 0.001 0.004 0.001 -0.196* -0.128  (0.09) (0.09) (0.09) (0.09) (0.09)       Former Communist 0.619*** 0.620*** 0.619*** 0.672*** 0.707***  (0.09) (0.09) (0.09) (0.10) (0.10)       Mediterranean 0.284* 0.287* 0.284* 0.402** 0.410**  (0.13) (0.13) (0.13) (0.13) (0.13)       187  1992 (ref: 2009) -0.155* -0.155* -0.155* -0.154* -0.153*  (0.07) (0.07) (0.07) (0.07) (0.07)       1999 -0.045 -0.044 -0.045 -0.044 -0.044  (0.06) (0.06) (0.06) (0.06) (0.06) Interactions      Working Class*Gini (ref: Salariat)  -1.790***   -1.420***   (0.47)   (0.36)       Intermediate Class*Gini  -1.114**   -0.787*   (0.37)   (0.33)       Working Class*GDP (ref: Salariat)   0.012      (0.01)         Intermediate Class*GDP   0.010      (0.01)         Working Class*Conservative    0.106 0.015 (ref: Salariat Liberal)    (0.06) (0.06)       Working Class*Social Democratic    0.360*** 0.237***     (0.06) (0.06)       Working Class*Former Communist    -0.106 -0.173***     (0.05) (0.05)       Working Class*Mediterranean    -0.208*** -0.220***     (0.06) (0.05)       Intermediate Class*Conservative    0.076 0.025     (0.05) (0.05)       188  Intermediate Class*Social Democratic    0.234*** 0.166**     (0.05) (0.05)       Intermediate Class*Former Communist    -0.045 -0.082*     (0.04) (0.04)       Intermediate Class*Mediterranean    -0.128** -0.136**     (0.05) (0.05)       Intercept 2.772*** 2.777*** 2.966*** 2.763*** 2.746***  (0.62) (0.62) (0.63) (0.62) (0.63) Random Effects      Country Wave .04*** .04*** .04*** .04*** .04***  (0.01) (0.01) (0.01) (0.01) (0.01) Class .01*** .01*** .01*** .00*** .00***  (0.00) (0.00) (0.00) (0.00) (0.00) Number of individuals 56834 56834 56834 56834 56834 Number of countries 24 24 24 24 24 Number of country-years 63 63 63 63 63 AIC 140817 140799 140820 140763 140750 Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001      

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0386719/manifest

Comment

Related Items