Alternative to Inglehart's Values Constructs Zoltán Lakatos Budapest University of Technology and Economics Working paper 6.488 words (excluding figures, tables, references, appendices, and abstract (153 words)) Abstract Despite abundant criticism regarding their theoretical foundations and composition, Inglehart's values constructs have become near currencies in cross-cultural research. Drawing on these critiques, this paper argues that Inglehart's three most influential indicators of cultural values: the Postmaterialism Index, the self-expression-survival, and the seculartraditional measures obfuscate the complexity of the value space at the individual level. To overcome the inherent limitations of Inglehart's instruments I present an alternative approach in the form of multiple correspondence analysis (MCA), a geometrical approach whose theoretical foundations are in field analysis. A recently developed variant of MCA performing rotation separates out those axes of cultural values (religiosity, authoritarianism, materialism, and achievement orientation) that Inglehart's scales treat as part of broader cultural dimensions. The results corroborate findings by other critics of Inglehart's methods suggesting that the failure to identify the libertarian-authoritarian axis as separate from both religiosity and materialism stems from a mechanistic view of the culture-economy dialectic. 1 An earlier version of this paper was presented at the R User conference in July 2012 in Bordeaux, France and the RC33 Conference of the International Sociological Association in Sydney, Australia in July, 2012. Acknowledgment Throughout this work I have benefited from collaboration with Marie Chavent and Michel van de Velden, developers of the software packages PCAmixdata for R and CAR for MatLAB respectively. Without their extensive support the statistical analyses applying a recently developed variant of multiple correspondence analysis would not have been possible. I also thank Róbert Tardos, Dominique Joye, Jörg Blasius, Nikosz Fokasz, Béla Janky, Zoltán Kmetty, János Balázs Kocsis, Ákos Kopper, and Frank T. Zsigo for their helpful comments and suggestions. I am solely responsible for any error and omission. Funding This work was supported by the Department of Sociology and Communication at the Budapest University of Technology and Economics (BME). 2 Postmaterialist Shift: Theses and Measures Empirical sociology's interest in cultural change goes back to the early history of social science. Over the past few decades, thanks to the evolution of information technology, representative surveys have been providing a growing body of cultural indicators aimed at measuring these processes. The World Values Survey (WVS), initiated by political scientist Ronald Inglehart stands out in thematic scope and geographical coverage among these programs. Drawing heavily on findings from the WVS, Inglehart's theses on and indicators of cultural change are of particular interest as they have profoundly influenced scholarly debates on the content of cultural values and their interconnections with various economic and political processes. Of Inglehart's three most influential measures, the Postmaterialism Index is a key indicator of the cultural shift presented in his early theses (Inglehart 1971, 1977) while the other two, the self-expression-survival and the secular-traditional values constructs (Inglehart 1997; Inglehart and Baker 2000) aim to tap a more complex set of phenomena. The latter two instruments reflect the increasing sophistication Inglehart's methodology has undergone, partly in response to the debates on his original theses. Nevertheless, as both the core of his theses and his approach to measuring values have remained constant, concerns with these three constructs are remarkably similar. Without undue simplification, the theoretical foundations of these measures comprise two complementary hypotheses. The scarcity thesis, borrowed from Maslow's needs theory (1970) posits that people value anything that is in short supply, therefore so long as physical survival and material security are not taken for granted, their value system is dominated by materialistic aspirations. An equally important aspect is the role attributed to religiosity and 3 authoritarianism, identified as "traditional values" and equally associated with economic scarcity and physical insecurity (Inglehart and Baker 2000). Increasing affluence leads to a de-emphasis of not only material gains but also religion and unquestioned authority in favor of secular, libertarian values, and "self-expression". The socialization thesis conceives of value orientations acquired during one's formative years as stable over adulthood. Periodical fluctuations, that is decreasing or increasing materialism within a given age cohort may occur as a result of mainly economic booms or downturns, but generations raised in significantly different material conditions retain their different values. Central to the empirical evidence behind the Postmaterialism Index is a factor analysis of responses to questions on national priorities (Table 1). Table 2 shows how each of the original 12 items load onto a single component. Given that most of these questions address policy preferences rather than values, the relationship posited by Inglehart between the private and the public domain is problematic. Flanagan (1982a, 1982b; Inglehart and Flanagan 1987) has argued that the Postmaterialism Index is an inadequate measure of two separate dimensions of political conflicts in Western societies: an emphasis on economic versus noneconomic issues and preference for libertarian, as opposed to authoritarian policies. The Postmaterialism Index classifies as "materialists" those who are simply authoritarians but see non-economic policy issues as paramount. The problematic inference of values from answers to questions on public policy is also discussed at length in Marsh (1975), Lafferty and Knutsen (1985), Trump (1991), Haller (2002), and Majima and Savage (2007). [Table 1 and Table 2 about here] Several authors have also pointed out that the evolution of the Postmaterialism Index does not follow the trajectory posited by Inglehart's theses. Growing portions of materialists 4 among the young in a number of European countries (Böltken and Jagodzinski 1985), the absence of significant effect of formative economic experiences on values in adulthood (Duch and Taylor 1993, 1994; De Graaf and Evans 1996; Clarke and Dutt 1991; Clarke, Dutt, and Rapkin 1997a, 1997b), and high instability, attributed specifically to random distribution of responses and non-attitudes (Van Deth 1983; Davis and Davenport 1999; Davis, Dowley, and Silver 1999) call into question the validity of the instrument, the underlying hypotheses or both. These problems are in part related to Inglehart's reliance on aggregated data for multivariate analyses. The factor analyses whose outputs he presents as evidence of a materialist-postmaterialist axis were performed on country averages. Inglehart justifies aggregation by the elimination of measurement error, arguing that the aggregate pattern is determined by those who "give «real» responses [because] random variations tend to cancel each other out" (Inglehart 1983: 89), an argument also advanced with regard to some of his later studies. However, aggregate consistence of factors and stability of indicators derived from aggregates may conceal inconsistence and instability at the individual level (Van Deth 1983; Davis et al.; Haller 2002; Majima and Savage 2007). This is a case of disaggregation error or ecological fallacy whereby individual attributes are inferred from macro indicators overlooking complexity at the micro level. With the WVS’s geographical coverage expanding and the availability of longitudinal data, Inglehart has constructed a more comprehensive series of indicators that he claims to account for much of the value change taking place across societies with very different cultural traditions (Inglehart 1997). The most important of his new measures are the traditionalsecular/rational and the survival-self-expression/well-being axes (Figure 1, Table 3) that define a two-dimensional value space in which societies form clusters that appear to represent historical-cultural entities in a "Global Cultural Map" (Figure 2). While the pool of variables used for this analysis is different from the policy preference questions that form the basis of 5 the Postmaterialism Index, their use is no less problematic. First, as with the Postmaterialism Index, Inglehart initially derived these scores from factor analysis performed on national aggregates, using country averages.1 Second, the composition of the resulting value dimensions exposes a reductionist approach similar to the construction of the Postmaterialism Index. Most importantly, both dimensions are too heterogeneous constructs to be considered basic value orientations. The traditional-secular/rational dimension ― in addition to the confusion about the labeling of its poles ― combines items tapping religiosity with indicators of authoritarianism and achievement values as if these were indicators of one single underlying dimension. The objection can be made ― and I will test the hypothesis below ― that this is not the case, i.e., the impact of non-authoritarian religiosity and authoritarian secularism is too important for religiosity and authoritarianism to be collapsed under one single construct. Of the variables from which the survival-well-being dimension were originally extracted, six ("happiness", "life satisfaction", "sate of health", "trust in people", "having [a sense of] free choice", and Affect Balance, an index of "positive versus negative" feelings) tap not values but psychological states. A seventh variable associated with this dimension is the Postmaterialism Index ("priority to economic and physical security over selfexpression and quality of life"), whose construction is problematic for the reasons reviewed above. It seems legitimate to suppose that had the psychographic variables and the Postmaterialism Index and the other non-values been omitted from Inglehart’s analysis, a much different configuration would have emerged from even a two-dimension solution. Obviously, Inglehart worked out the cultural map to remedy the major flaws of the Postmaterialism Index, partly in response to criticism of his postmaterialism theses. Still, he seems to have followed the same path by treating preferences as indicators of deep-seated 1 This remark concerns the first version of the two-dimensional cultural map (Inglehart 1997: 82) but the same issues persist with regard to the second version (Inglehart and Baker 2000), a reduced, 10-item battery backed by a factor analysis at the individual as well as the aggregate level. See remarks for Table 3. 6 orientations, combining qualitatively distinct orientations into one single dimension (libertarian and economic concerns under the postmaterialist label and later under "selfexpression", authoritarianism and religiosity under "traditional values"). [Figure 1, Table 3 and Figure 2 about here] Finally, it is remarkable that equivalence issues rarely come up in the debates on Inglehart's measures (Haller 2002). One such exception is a study by van De Vijver and Poortinga (2002) in which the authors tested the equivalence of the Postmaterialism Index and found that it is configurally invariant at the individual level in all but one of the 39 regions examined but added that the construct becomes more salient in affluent countries. However, they also concluded that in order to be comparable across countries, some items have to be eliminated. Such issues may eventually compromise the comparability of the secular-traditional and the self-expression-survival scales, although findings of a comprehensive test (involving not only configural, but also metric and scalar invariance) are yet to be made available.2 Multiple Correspondence Analysis A sensible alternative to Inglehart's instruments has to, at a minimum (1) build on items that are actual values, (2) focus on the individual level, and (3) avoid ambiguity in the resulting constructs. In addition, (4) it should restrict cross-cultural comparisons to (country) scores satisfying scalar invariance. The choice of multiple correspondence analysis (MCA) 2 In comparison, the Schwartz value scales have been subjected to a comprehensive test of invariance and were found to satisfy configural and metric (but not full) equivalence in 20 countries (Davidov, Schmidt, and Schwartz 2008). 7 for the study presented in this section follows from the first three considerations. Given that equivalence testing of summary indicators provided by MCA is still mostly unexplored territory, the fourth requirement is a direction for future research. Although its empirical potential is amply demonstrated thanks notably to Bourdieu's work on taste and cultural stratification (1979), correspondence analysis is still rarely applied in sociological studies. A geometric data analysis (GDA) method using an observationsproperties table, MCA does not impose any constraint on the data (Blasius and Greenacre 2006), unlike factor analysis which assumes linear constraints. MCA does not require the averages for the variables whose structure it seeks to reveal and thus allows the researcher more freedom in identifying latent structures based on how categories of the analyzed variables are plotted in an n-dimensional space. Given that many of Inglehart's factor analyses are performed on national aggregates, linear constraint was not a concern in his studies but since his variables are categorical, scale properties become an issue once his factor analyses are performed at the individual level.3 Another consideration in favor of MCA is its focus on relations rather than variables: instead of looking for combinations of independent variables that might explain a priori dependent variables, "MCA starts from and carefully constructs that which is to be explained" (Le Roux et al. 2008: 1053). The power of correspondence analysis lies in its representation of that complexity at two levels: a geometrical space defined by the relations between the properties (in this case values) and a space where the units of observations (individuals) are located (Rouanet, Ackermann, and Le Roux 2000). The closer individuals are positioned within that space, the more properties they share, and inversely. This mirrors Bourdieu's meta-theoretical framework which "identifies reality not with 3 In addition, many of Inglehart's items have dissimilar scales, from binary to 4- or 10-point Likert-type, etc. Such scale inconsistencies render the application of multivariate methods using linear methods problematic. 8 substances but with relations" (Bourdieu 1987: 150, emphasis added), a relational way of thinking that Bourdieu traces back to the origins of modern science.4 The substantialist framework is evident in many of Inglehart's writings, most significantly in his adoption of a methodological apparatus that is reminiscent of the "sociology of variables", a perspective that Bourdieu repeatedly criticized for its mechanistic approach and reductionism (Rouanet et al. 2008). The focus on the economic "origins" of postmaterialism and other values, laid out in Inglehart's scarcity and socialization hypotheses as well as their later, refined versions (e.g., the emphasis on "individual resources" in the theory of human development, (Welzel, Inglehart, and Kligemann 2003; Welzel 2002) is one example of that deductive logic. Another is the opposite perspective, also adopted by Inglehart in his later work on the macroeconomic "effects" of cultural values (Granato, Inglehart, and Leblang 1996a, 1996b) where he and his coauthors argue that based on the endogenous growth model proposed by economists (e.g., Lucas 1988; Romer 1990) to overcome the limitations of neoclassical theories of economic growth (Solow 1956; Swan 1956), a cultural or value component is assumed to have an independent impact on productivity and thus to contribute to economic growth. The latter argument is presented with reference to the dominant interpretation of Weber's Protestant ethic thesis (Weber, Baehr, and Wells 2002) which, according to Granato et al. suggests an independent effect of culture on economic outcomes. However, as recent scholarship on the Protestant ethic thesis has shown, Weber's views on the interconnections between cultural and economic phenomena are not what is generally assumed by their dominant interpretations as Weber never proposed a theory of "causation" by cultural values. Instead, he conceived of the relationship between culture and the economy as "elective affinities" (Howe 1978; Chalcraft and Harrington 2001; 4 For a general discussion of the bourdieusian conception of social space with implications for CA, see Wacquant (2013). 9 Löwy 2004; Chalcraft 2005; Runciman 2005; Treviño 2005; McKinnon 2010), an insight that informed Bourdieu's concept of field and habitus (Bourdieu 1980: 34). In short, Inglehart's scholarship treats values once as outcomes, once as causes of macro (mainly economic) phenomena. However, none of these accounts grasps the complexity of how values are embedded in the socio-economic context. Social contexts are fields where structure and agency, material and symbolic, economic and cultural are inextricably bound together in ways irreducible to a "dominant effect" or causality. Values share with Bourdieu's habitus the property of being a modus operandi, a generating formula of action, a system of stable dispositions that is both structured and structuring (Bourdieu 1980: 88). The space of values is a field where agents occupy different positions according to their own "normative emphases that underlie and justify the functioning of social institutions" (Schwartz 2011: 314, emphasis added) ― where the terms "underlie" and "justify" capture both the materiality and the symbolic aspects of values "put to work". Such a perspective "translates the question of 'why' into a 'how' that is answerable in terms of a situation and its typal vocabulary of motives" (Mills 1940: 906). The theoretical framework in which correspondence analysis finds its empirical utility is therefore a departure from both the rigid structuralist and the substantialist methods of inquiry. MCA is thus the method of choice in field analysis. Overcoming the limitations of competing theoretical traditions around the "rather formulaic fractal divide between the macro and micro, society and individual, material and social" (Savage and Silva 2013: 124), field analysis ― and by implication, MCA ― concentrate on that which the sociology of variables ignores: relations.5 In addition to Bourdieu's wide-ranging applications (besides his now classic "Distinction" (1979), these include Bourdieu and De Saint Martin 1978; De Saint Martin and Bourdieu 1987; Bourdieu 1999, 2000), the empirical relevance of MCA to the actor-field 5 For a discussion of field theory in the social sciences, see Martin 2003. 10 approach is supported by a number of recent studies on taste and lifestyles (Le Roux et al. 2008; Silva and Wright 2009; Blasius and Mühlichen 2010; Bennett, Bustamante, and Frow 2013; Silva and Le Roux 2011; Hanquinet 2013; Hanquinet, Roose, and Savage 2013; Purhonen and Wright 2013), fields of power (Hjellbrekke et al. 2007; Lebaron and Le Roux 2011; Flemmen 2012; Buhlmann, David, and Mach 2013; Kropp 2013), class and social inequalities (Veenstra 2007), and political attitudes (Harrits et al. 2010). Most relevant to our subject however is a study on the British value system by Majima and Savage (2007) that, in addition to applying MCA because it provides a representation of the complexity of "individual value spaces" is also a critical engagement with Inglehart's work. Crucially from our point of view, Majima and Savage conclude that "the supposed shift towards postmaterialist values that Inglehart claims for the British case, as for other countries, is in fact an artefact of questionnaire design" (ibid: 312). In the following I extend on their analysis. The Cloud of Modalities Unrotated Solution I use data from the first four waves of the WVS, collected between 1981 and 2004 in 84 countries, a total sample of 267.870 (World Values Survey Association 2009). The choice of variables is an aspect where constraints follow from the fact that only a smaller set of questions were asked in all four waves. The 28 questions selected and combined into 19 variables that were used in the analysis reflect a compromise between the requirements stemming from the criticisms reviewed above (i.e., the questions should tap values, not policy preferences or mental states), and the availability of variables across four waves. The original 11 responses were recoded into binary modalities (see Table 5 in the Appendix for recoding rules).6 [Figure 3 about here] Figure 3 shows the first four axes emerging from the unrotated solution of our 19 variables. At first inspection, the cloud of modalities defining the first two axes looks similar to Inglehart's 2-dimensional value space presented in the Global Cultural Map in Figure 2. The first dimension (horizontal in upper plane) translates an opposition between values associated with traditional morality (on the right) and their opposite (on the left). Traditional morality consists of strong religiosity and authoritarianism which stand in opposition to secular and libertarian values. The second (vertical) axis represents an opposition between achievement orientation (bottom) and materialistic attitudes (top).7 In addition, modalities with higher than average contribution to Axis 2 include some items associated with authoritarianism and libertarianism. Axis 2 therefore looks remarkably similar to Inglehart's self-expression-survival scale. The third dimension (horizontal, in bottom plane) captures an opposition between on one hand an attitude that is both achievement-oriented and materialistic8 and on the other its opposite, an attitude that values neither achievement nor material gain. This latter pole also includes a rejection of outgroups. The fourth (vertical) axis translates a conflict between an achievement orientation that is not materialistic and a materialistic outlook lacking any sense of achievement. 6 The only exception is the frequency of participating in religious services, recoded in three categories. For ease of interpretation, the maps do not show those modalities whose contribution to the axes is lower than the average, defined as dividing 100 by the total of modalities. The contributions of the modalities are included in Table 4 in the Appendix. 8 As shown by the high contribution of the modalities "achievement most important" and "good pay most important". 7 12 The semantic heterogeneity of the four axes (religiosity combined with authoritarianism, materialism with achievement values, etc.) makes this rendering of the value space difficult to interpret. At the same time, alternative axes emerge in both diagrams. With Plane 3 and 4, an alternative axis running from the top right to the bottom left quadrant is a more straightforward opposition because it simply reflects the importance attached to achievement without the interference of the items related to materialism. Likewise, another line linking the top left to the bottom right quadrant is a better representation of materialism than Axes 3 or 4. The same opportunity for simplification is apparent with regard to Axes 1 and 2. A line running from the bottom right to the top left quadrant captures religiosity without the influence of the modalities related to authoritarianism. And vice versa, linking the bottom left and the top right quadrants is another line tapping authoritarianism without religiosity. These distributions suggest that the rotation of the axes would yield semantically more homogeneous dimensions. In "MCA as we know it" such alternative axes often make more sense than the original ones, yet these cannot be provided in most current statistical packages.9 This is not necessarily an inconvenience when summary indicators are not sought ― but becomes a hindrance once the aim of the study is to also provide such indicators. Rotated Solution Rotation in multiple correspondence analysis has been proposed by a number of authors (Kiers 1991; Adachi 2004; Velden and Kiers 2005). Kiers's idea was to extend the varimax criterion to arrive at simple structures in PCAMIX, which is a principal component method for a mixture of qualitative and quantitative variables. The varimax criterion is expressed 9 Discussing the configuration of modalities in the 3-axis solution of their MCA of 21 variables from the WVS, Majima and Savage also conclude that "rather than (…) endorsing the centrality of the materialist-post-materialist (or 'survival-selfexpressive') dimension, it [i.e., their MCA solution] suggests ruptures between libertarian and authoritarian views" (2007: 305). Nevertheless, their MCA only provides an unrotated solution where the corresponding items emerge counter-posed along a virtual axis (linking the top-right and the bottom-left quadrants, see Figure 2 in Majima and Savage 2007: 306). 13 with squared loadings defined as correlation ratios for qualitative and squared correlations for quantitative variables. Later formulations (Saracco, Chavent, and Kuentz 2010) use a singular value decomposition (SVD) approach to produce the component scores, squared loadings and the principal coordinates of the categories. This procedure searches for an optimal rotation using the iterative procedure suggested by Kaiser (1958) for principal component analysis. It leads to the same rotated solution as Kiers’ method, but is computationally more efficient. Until very recently, these techniques were not available in statistical software. The first available are CAR (Correspondence Analysis with Rotation), under MATLAB developed by Lorenzo-Seva, van de Velden, and Kiers (2009), and PCAMIXdata, a package for R by Saracco et al. (2010). The rotated solutions presented in this section were obtained with these two packages.10 [Figure 4 about here] We saw that authoritarianism and religiosity one hand, and achievement orientation and materialism on the other would not separate out if we were to retain the axes provided by the unrotated MCA solution. In this respect, the unrotated solution is reminiscent of Inglehart's factor analyses where these domains are subsumed under the more comprehensive seculartraditional and self-expression-survival constructs. However in the rotated 4-axis solution presented in Figure 4, the new axes are aligned precisely along those lines that appear as a sensible alternative to the original (unrotated) axes. Here religiosity, authoritarianism, materialism and achievement values (Axes 1, 2, 3, and 4, respectively) constitute unambiguously separate dimensions. Compared with the unrotated solution, the percentage of 10 The two packages were used complementarily. The exploratory analyses were performed using CAR, based on the summary Burt table. CAR also calculated the contribution of modalities to the axes. PCAmixdata provided the coordinates at the individual level for the retained unrotated and rotated solutions. The results presented below come from an MCA with orthogonal rotation. 14 variance (inertia) explained by the first dimension dropped significantly (from 27.1 to 16.3%) while that of the second dimension increased (from 6.4 to 9.9%). There were also increases in the variances explained by the third and the fourth dimensions (from 4.9 to 7.8%, and from 3.1 to 7.6%). This redistribution of the variances indicates that the axes resulting from the rotated solution are semantically more coherent than the more ambiguous scales in the unrotated solution. The most spectacular gain in consistence is due to the picking up of parts of the inertia from the first unrotated dimension by Axes 2 to 4 in the rotated solution. Also, while items related to materialism and achievement appear as items with significant contribution also in Plane 1-2 of the unrotated solution, they are of no significance to the first two dimensions in the rotated solution, religiosity and materialism. The same can be said of a couple of items related to authoritarianism and libertarianism (e.g., acceptance and rejection of divorce and single mothers) that in the unrotated solution have significant contributions to Axes 3 and 4 but are not relevant to the third and fourth dimensions in the rotated solution, capturing materialism and achievement values in a more straightforward fashion. The resulting increase in homogeneity has led to scales that are easier to interpret and thus provide a more plausible description of the value space than the semantically less consistent axes in Inglehart's Global Cultural Map. We saw that Inglehart conceives of materialism as part of a broader domain that includes also "traditional" and "survival" values. The emergence of materialism as a compact construct in the rotated 4-axis substantiates the various criticisms addressed to the postmaterialism theses, reviewed above. One could object that Inglehart's typology is based on a minimalist configuration and other dimensions would emerge were we to carry his analysis beyond just two dimensions. Indeed, Inglehart readily admits that we would get a "somewhat different solution" if further dimensions were retained in his factor analyses. Still, he dismisses that suggestion as misleading by arguing that "(d)oing so produces a far more 15 complicated result that might superficially seem more scholarly" whereas "cross-cultural variation (…) can be interpreted with a relatively parsimonious model" (Inglehart 1997: 91). However, when techniques adapted to categorical data and the complexity of the individual level such as MCA are applied carefully, the result is not more but less complicated latent constructs. If anything, it is his constructs that are unnecessarily complex because they obfuscate rather than clarify the structure of values and the direction of cultural change. The breaking down of the 19 variables into four semantically distinct dimensions at the individual level is an indication of substantive differences in the content of these constructs, rather than an artifact of "scholarly sophistication". The axes resulting from the rotated MCA solution translate values that are incommensurable and thus should be kept separate. They stand as empirical support to the core of Inglehart criticisms. Inglehart's summary indicators are indeed reductionist: religiosity is not a corollary of authoritarianism, it is therefore wrong to combine these two distinct values in a secular-traditional construct. A sizeable portion of the sample is religious and libertarian or secular and authoritarian ― dispositions that the two-dimensional Global Cultural Map fails to discern. Likewise, materialism differs from and should not be conflated with authoritarianism, religiosity, and low priority to achievement. Whereas in Inglehart's typology, materialism entails a complex set of attitudes, actual materialism, as measured by Axis 3 in the rotated solution is (really) only about the importance of material gain and nothing else ― and its opposite value is not post- but non-materialism. [Table 4 about here] 16 The Cloud of Individuals Given the scarcity of scholarship on the techniques exploring the equivalence of constructs resulting from MCA and related methods,11 any comparison of scores produced by the solutions presented in the previous section has to be treated with caution. For this reason, this paper leaves the discussion of country scores to further studies, probably until advances in the comparability of MCA solutions are made. With that qualification, an overview rather than detailed country comparisons may reasonably complement the findings related to the cloud of modalities. Landmark Countries and Individuals Studies using MCA sometimes describe landmark individuals in order to make better sense of the clouds (e.g., Le Roux et al. 2008; Purhonen and Wright 2013). The four landmark cases presented in Figure 5 have been selected to represent "typical" individuals from four national samples that stand also as "typical" in the sense that the distribution of respondents in their respective clouds is illustrative of easy-to-interpret value configurations. This is because the majority in these clouds are clustered further away from the origin than in any other national samples. "Typical" here refers to ideal types in the Weberian sense, i.e., an accentuation of certain characteristics (Hendricks and Breckinridge Peters 1973) of more or less homogeneous clusters of national samples and should be understood as "caricatures" rather than country medians. The Jordanian sample (collected in 2001) is characteristic of a premodern cultural profile: mostly religious, authoritarian, non-materialist and not achievement-oriented. The majority in the Russian sample (1995) display a value system typical of societies marked by several decades of communist rule: secular, authoritarian, materialist and not achievement11 The few exceptions include Blasius and Thiessen (2006a, 2006b) and Blasius and Graeff (2009). 17 oriented. The Swedish sample (1999) is a showcase of late modern societies: mostly secular and libertarian, while also achievement-oriented but non-materialistic. The Colombian sample (1998) is for the most part religious, libertarian, but also materialistic and achievement-oriented. In addition to being ideal types, these four national clouds are significant for a further consideration: two of them represent configurations that do not fit Inglehart's typology. The Jordanian sample, while authoritarian, is not materialistic, and the majority of Colombians, while strongly religious, are not authoritarian. Only the Swedish and the Russian clouds conform to the combinations "allowed" by Inglehart's Global Cultural Map: secular and libertarian, non-materialist and achievement oriented, and vice versa. As their respective positions in the maps suggest, the landmark cases correspond to individuals with different cultural values. More interesting, however, are their similarities and differences regarding variables that are not included in the clouds of modalities as these are instructive of the maps’ explanatory potential. The following description focuses on this latter category. [Figure 5 about here] Secular and authoritarian [top-left], materialist and not achievement-oriented [bottom-right] Case 122 955, "Svetlana" is a Russian woman aged 70 (surveyed in 1995) living alone in a big city in Central Russia. She has no sense of freedom and control in her life and is generally distrustful of others. Very dissatisfied with the people in national office, she thinks that Russia is run by a few big interests. She finds that compared to 10 years before (i.e., under communism), a larger share of Russians live in poverty, which she attributes to unfair conditions in her country. For Svetlana, a democratic system is fundamentally bad ― probably because of her own experiences of what goes under the label of democracy, rather 18 than out of a rejection of democracy itself ―, a stance that explains her preference for military rule and a strong leader. Overall, she says that Russia's political system was better under the communist regime, and would vote for a small nationalist party. Secular and libertarian [bottom-left], non-materialist and achievement-oriented [topleft] Case 189 046, "Magnus" is a married man aged 61 (surveyed in 1999), living in a household of two in a medium-size city in Southern Sweden. In the past he formally belonged to a Protestant denomination, but qualifies himself as "not religious". He is very satisfied with his life, albeit with a limited sense of freedom and choice. Generally trustful of others, Magnus is a member in several organizations and associations, and has done voluntary unpaid work. He thinks that people, rather than the government should take more responsibility to ensure their own well-being and views competition as good, rather than bad. At the same time, he favors environmental legislation, and even agrees with tax increases to prevent environmental pollution. Regarding the political system, Magnus rates Sweden’s as rather good, and would vote for a centrist liberal party. Religious and libertarian [bottom-right], materialist and achievement-oriented [topright] Case 104 551, "Isabel" is a 25-year old (surveyed in 1998) childless single woman living with her parents in a Central Colombian city. She is both very religious and antiauthoritarian, an attitude that is not in short supply among Colombians. She is accepting toward childless women, divorce and outgroups, and has participated in subversive political action. Satisfied with her financial situation as well as with her life in general, she feels a great deal of freedom and control but is distrustful of others. Regarding attitudes toward 19 government and politics, Isabel's support for Unión Patriótica, a socialist party is probably related to her assessment that Colombia is run by a few big interests and poverty is on the rise as well as her preference for increased government ownership. However, Isabel’s preference for radical change is confined within democratic limits as she rejects alternatives by the military or a strong man. Religious and authoritarian [top-right], non-materialist and not achievement-oriented [bottom-left] Case 239 119, "Rashid" is a 28-year old single, childless man (born 1973) living with his parents in a small town in North-West Jordan. He has completed high school and is employed as a non-manual office worker. Rashid attends mosque several times a week and is for the unconditional respect of parents, and thinks that men make better political leaders than women do. He defines himself as a Muslim above all (i.e., before being Jordanian or Arab, etc.), and has also a strong clan identity. Satisfied with his life, he is totally disengaged from politics, would not vote for any party and prefers people with strong religious beliefs in public office. Discussing landmark individuals from two separate MCAs resulting from a British and a Finnish survey on cultural taste, Purhonen and Wright note that "(e)ven as we cannot claim that any one individual is typical or representative of their position, nor can we dismiss the meaningfulness of their account of that position as some sort of empirical reality" (2013: 265). The same proposition is warranted here with the addition that while the properties discussed above that were not included in the MCA obviously cannot be fully accounted for by the positions that the four landmark individuals occupy in the clouds, many of them "make sense" in light of the corresponding individual values. It would be a mistake to dismiss such 20 similarities and dissimilarities as merely anecdotal. The positions occupied by individuals in the clouds are meaningful beyond the variables used in the MCA. Relationship with macro variables The rationale for better measures is to also better understand how cultural values are related to various macro phenomena. Because the postmaterialism thesis posits that materialism and "traditional values" are negatively correlated with national affluence, comparisons with national income seem an obvious choice.12 (As we have seen, it posits a causal relationship, national affluence being an independent, and values dependent variables.) Inglehart stresses that it is national rather than individual or household affluence that gives rise to postmaterialism (Inglehart and Abramson 1994). Comparisons between his Postmaterialism index and per capita GDP are often included in support of the proposition that from poor to rich countries, there is an increase in the portion of postmaterialists. This relationship is captured in Figure 6 which shows a strong negative linear relationship between per capita GDP and the Postmaterialism Index (upper left diagram). Inglehart's index steadily decreases from the poorest to the richest nations. Using our materialism scores from the rotated MCA solution, the picture is more complex, suggesting a curvilinear, rather than linear relationship (upper right diagram in Figure 6). In contrast with the high scores of Inglehart's Index, our new materialism score is very low among the poorest countries. It then sharply increases in the 3 to 10 thousand dollars per capita GDP range before dropping again below zero. But this is not to say that we have identified a mechanism whereby small increments in overall affluence first give rise to an increase, and, at much higher levels to a decrease in materialistic aspirations. That is because most societies in the lower-middle per 12 As with the cloud of individuals, these associations need to be treated with caution until the equivalence of constructs is established. For this reason, regressions are not included in this section. 21 capita GDP range (where our materialism score reaches its highest values) have lived, until the early 1990s under authoritarian regimes, and nearly all positive materialism scores belong to these post-authoritarian regimes (marked in different data points, postcommunist countries for the most part). Their higher materialism scores suggest that materialism has probably less to do with overall affluence than other contextual variables. In the case of societies under post-authoritarian regimes, most of their data come from the early phase of their democratization marked by rapid privatization, rising social inequalities and unemployment, and the worship of entrepreneurship and material success by the new elites and their official ideologues. We cannot at this level determine the extent to which these or other factors contribute to the higher materialism of the lower middle per capita GDP range but can reasonably suggest that overall affluence is not a major factor. [Figure 6 about here] What actually decreases from poor to rich countries is authoritarianism (second diagram on the right in Figure 6), giving further support to the already corroborated claim that Inglehart's Postmaterialism Index taps the libertarian-authoritarian cleavage rather than (post)materialism. Moreover, per capita GDP's negative correlation with authoritarianism is a great deal stronger than with Inglehart's traditional-secular measure (bottom left diagram) because the latter is composed of items related to religiosity as well as authoritarianism. From poor to rich countries, the decrease in religiosity is much less pronounced (bottom right diagram) than the drop in authoritarianism. The above comparisons suggest that the relationship between values and macroeconomic conditions, specifically national affluence is not in line with the postmaterialism thesis and its offshoots. With the exception of authoritarianism, differences 22 in per capita GDP correspond to only small differences in values. With regard to materialism, even small differences appear to be related to other variables ― most importantly, the historical context ― rather than national affluence. Historical context might also explain away a great deal of the correlation between national affluence and authoritarianism. Were we to use Inglehart's instruments, these particularities would not come to light. These findings support a key insight in field analysis which, by virtue of its focus on relations stresses the stability of social configurations (Martin 2003). Such stability is at odds with the mechanistic view prevailing in the sociology of variables where variations in some "ultimate" factors induce profound changes in the social context. Conclusion The results presented in this article confirm what many Inglehart critics have long been suggesting. Rather than being a showcase of "scholarly sophistication", instruments adapted to the complexities of the individual level are a requirement to better grasp the content of values and the direction of cultural change. The apparent efficiency of metric methods applied to aggregate data comes at the price of reductionist typologies and indicators. Specifically, religiosity, authoritarianism, and materialism constitute distinct value domains that should not be subsumed under broader constructs. Overlooking the complexities of individual value configurations has also led to misconceptions about their "origins" on one hand and "consequences" on the other, sometimes in the form of contradictory findings even within the same school of research. Inglehart's work on values is a demonstration of this struggle where the methodological apparatus follows from misplaced theoretical perspectives. Hence the reorientation on the 23 actor-field dialectic called for in value research also requires that the culture-economy dichotomy be abandoned as empirically irrelevant. 24 Tables and Figures Table 1 The 12-item battery of materialist and postmaterialist measures Materialist Measures Postmaterialist Measures A) Maintaining order in the nation* B) Giving people more say in decisions on the government* C) Fighting rising prices* D) Protecting freedom of speech* E) Maintaining a high rate of economic growth G) Giving people more say in how things are decided at work and in their community F) Making sure the country has strong defense forces H) Trying to make our cities and countryside more beautiful I) Maintaining a stable economy K) Moving towards a friendlier, less impersonal society J) Fighting against crime L) Moving towards s society where ideas count more than money *Items used for the 4-item materialist/postmaterialist battery. Reproduced from Inglehart 1997: 355. The Postmaterialism Index is calculated using a 4- or 12-item battery pertaining to national priorities and policy preferences as perceived by the respondent. The 4-item index is constructed as follows: for the questions on the first and the second most important national priorities, respondents selecting both "maintaining order in the nation" (A) and "fighting rising prices" (C) are classified as materialists, while those selecting both "giving people more say in decisions on the government" (B) and "protecting freedom of speech" (D) are classified as postmaterialists. Those selecting both a "materialist" and a "postmaterialist" item are classified as mixed. This 3-point scale is available for most WVS country surveys, whereas the 12-item index (on a 6-point scale, constructed from questions A to L in three different sets on policy goals) is only available for a limited number of countries and waves. Table 2 The Materialist/Postmaterialist Dimension in 15 Western Nations, 1990 (Loadings on First Principal Component in Factor Analysis) Item K) Less Impersonal Society G) More Say On Job B) More Say in Government L) Ideas Count More than Money D) Freedom of Speech H) More Beautiful Cities Loading 0.60 0.62 0.50 0.51 0.34 0.18 F) Strong Defense Forces C) Fight Rising Prices J) Fight against Crime A) Maintain Order E) Economic Growth I) Maintain Stable Economy -0,26 -0,28 -0,39 -0,55 -0,59 -0,63 Items with materialist polarity are in boldface type; items with Postmaterialist polarity are in italics. Reproduced from Inglehart 1997: 112, table 4.1 (codes referring to Table 1 added), based on 1990-91 World Values Survey data from France, Britain, West Germany, Italy, the Netherlands, Denmark, Belgium, Ireland, Norway, Sweden, Denmark, Finland, Iceland, Canada, and the United States. 25 Figure 1 Inglehart’s two-dimensional cultural typology: “Traditional v. secular/rational” and “Survival v. wellbeing” values Secular/Rational Authority Survival Well-Being Traditional Authority Reproduced from Inglehart 1997: 82, Figure 3.2. Data are from the 1990-93 wave of the World Values Survey. The figure shows the first and the second principal components (horizontal and vertical axes respectively, accounting for 30 and 21% of the variance) emerging from a factor analysis of data from representative national surveys of 43 societies, aggregated to the national level. The positioning of the items reflects their loadings on the two dimensions. 26 Table 3 Items Characterizing Two Dimensions of Cross-Cultural Variations Traditional v. Secular-Rational Valuesa Survival v. Self-Expression Valuesc TRADITIONAL VALUES EMPHASIZE THE FOLLOWING: SURVIVAL VALUES EMPHASIZE THE FOLLOWING: God is very important in respondent’s life Respondent gives priority to economic and physical security over self-expression and quality-of-lifed It is more important for a child to learn obedience and religious faith than independence and determinationb Respondent describes self as not very happy Respondent has not signed and would not sign a petition Abortion is never justifiable Respondent has strong sense of national pride Homosexuality is never justifiable Respondent favors more respect for authority You have to be very careful about trusting people (SECULAR-RATIONAL VALUES EMPHASIZE THE OPPOSITE) (SELF-EXPRESSION VALUES EMPHASIZE THE OPPOSITE) Principal components from a factor analysis Inglehart-Baker (2000: 24) without the factor loadings present in the original table. Nation-level and individual-level data from 65 societies surveyed in the 1990-91 and 1995-98 World Values Surveys. Compared with the analysis shown in Figure 1, this is a reduced, 10-item battery. The Survival v. Self-Expression scale is identical with the Survival v. Well-Being scale in Figure 1. a Explains 44 percent of cross-national variation and 26 percent of individual variation b Autonomy index c Explains 26 percent of cross-national variation and 13 percent of individual variation d Measured by the four-item materialist/postmaterialist values index. 27 Figure 2 Global Cultural Map: Locations of 65 Societies on Two Dimensions of Cross-Cultural Variation: World Values Surveys, 1990–1991 and 1995–1998 Reproduced from Inglehart-Baker 2000: 29, Figure 1. The scales on each axis indicate the country’s factor scores on the given dimension. The positions of Colombia and Pakistan are estimated from incomplete data. 28 Figure 3 Cloud of modalities from unrotated MCA solution in plane 1-2 (upper map) and 3-4 (lower map) -0.7 -0.2 0.3 Survival (+) -1.2 0.8 0.8 THRIFT God not imp. no comfort from religion REJECTS OUTGROUPS NO SUBVERSIVE ACTION no life after death CHILDLESS WOMAN NOT 0.3 OK divorce not OK ACHIEVEMENT service: never Axis 2 [λ2=0.064] FIGHTING RISING PRICES MOST IMP . GOOD PAY MOST IMP. NOT IMP. single mother not OK learn faith not imp. God imp. divorce OK NO THRIFT comfort from religion learn faith imp. life after death GOOD PAY NOT IMP. -0.2 service: 1+/month SUBVERSIVE ACTION (-) Self-expression CHILDLESS WOMAN OK RESPECT PARENTS CONDITIONAL IMAGINATION -0.7 ACHIEVEMENT MOST IMP . -1.2 (-) Secular Axis 1 [λ1=0.271] Traditional (+) Multiple correspondence analysis of 19 variables from four waves of World Values Survey data, 1981-2004. See Table 5 in the Appendix for recoding rules and Table 4 for contributions to the axes. Only modalities with a contribution exceeding the average as defined by dividing 100 by the total number of modalities (100/39 = 2.56%) are shown. Modalities with significant contributions are in lower case for Axis 1 and 3, SMALL CAPITALS for Axis 2 and 4, and italics for more than one axis. The combined contributions of these modalities to Axes 1 to 4 are 86, 76, 89, and 83% respectively. Non-mater. & Achiev. imp.? (+) -0.8 -0.3 0.2 ACHIEVEMENT MOST IMP. 0.7 SERVICE: NEVER SINGLE MOTHER rejects outgroups NOT OK Axis 4 [λ4=0.031] GOOD PAY NOT IMP . (-) Mater. & Achiev. not imp.? 0.7 DIVORCE NOT OK work/life balance not imp. 0.2 interesting job interesting job not imp. ACHIEVEMENT work/life balance most imp. NOT IMP. -0.3 DIVORCE OK SERVICE: OCCASIONAL SINGLE MOTHER OK FIGHTING RISING PRICES MOST IMP . GOOD PAY MOST IMP . -0.8 (-) Non-mater. & Achiev. not imp.? Axis 3 [λ3 =0.049] 29 Materialist & Achiev. imp.? (+) Figure 4 Cloud of modalities from rotated MCA solution in plane 1-2 (upper map) and 3-4 (lower map) -0.7 -0.2 0.3 Authoritarian (+) -1.2 0.9 REJECT OUTGROUPS DIVORCE NOT OK SINGLE MOTHER NOT OK NO SUBVERSIVE ACTION OBEDIENCE men better learn faith imp. political leaders abortion not OK THRIFT CHILDLESS WOMAN NOT Axis 2 [λ2 =0.099] 0.8 service: never 3+ children denomination God imp. God not imp. no denomination learn faith not imp. no comfort from religion less respect for authority 0.4 service: 1+/month euthanasia not OK prayer outside service never think about meaning of life no life after death OK comfort from religion life after death good & evil not clear -0.1 independence marriage outdated sexual freedom euthanasia OK SUBVERSIVE ACTION abortion OK SINGLE MOTHER OK CHILDLESS WOMAN (-) Libertarian suicide OK prostitution OK DIVORCE -0.6 OK IMAGINATION OK RESPECT PARENTS CONDITIONAL -1.1 (-) Secular Axis 1 [λ1=0.163] Religious (+) Multiple correspondence analysis of 19 variables from four waves of World Values Survey data, 1981-2004. See Table 5 in the Appendix for recoding rules and Table 4 for contribution to the axes. Only modalities with a contribution exceeding the average as defined by dividing 100 by the total number of modalities (100/39 = 2.56%) are shown. Modalities with significant contributions are in lower case for Axis 1 (religiosity) and 3 (materialism), SMALL CAPITALS for Axis 2 (authoritarianism) and 4 (achievement), and italics for more than one axis. The combined contributions of these modalities to Axes 1 to 4 are 89, 75, 82, and 84% respectively. Modalities in small lighter fonts are categories of supplementary variables that do not contribute to the axes and are plotted as an illustration. -0.8 -0.3 0.7 1.2 1.3 Achievement imp. (+) Axis 4 [λ4=0.076] 0.2 ACHIEVEMENT MOST IMP . WORK /LIFE BALANCE NOT IMP. doing an important job most imp. 0.8 interesting job not imp. good pay not imp. 0.3 job useful for society not imp. (-) Achievement not imp. imagination job useful for society imp. -0.2 interesting job THRIFT SERVICE: OCCASIONAL good pay most imp. material benefits most imp. reject outgroups WORK /LIFE BALANCE MOST IMP . ACHIEVEMENT NOT IMP. FIGHTING RISING PRICES MOST IMP . -0.7 (-) Non-materialist Axis 3 [λ3 =0.078] 30 Materialist (+) Table 4 Contribution of modalities to dimensions 1 to 4 in the unrotated and the rotated MCA solutions Unrotated solution dim 1 Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Modality dim 2 Label Traditional Survival Rotated solution dim 3 dim 4 dim 1 dim 2 dim 3 dim 4 Materialism & Achievement Nonmaterialism & Achievement Religiosity Authoritarianism Materialism Achievement 1 respect parents conditional 3.4 4.3 0.3 0.8 1.1 7.3 1.0 0.3 2 always respect parents 1.1 1.1 0.1 0.2 0.4 2.0 0.2 0.1 1 learn faith not imp. 4.2 0.3 0.0 0.0 4.1 0.9 0.2 0.0 2 learn faith imp. 8.7 0.6 0.0 0.1 8.3 2.1 0.3 0.1 1 no imagination 0.2 1.3 0.5 0.0 0.0 1.0 1.0 0.0 2 imagination 0.8 5.4 1.9 0.0 0.1 3.9 4.2 0.2 1 no obedience 0.9 1.0 0.0 0.1 0.4 1.6 0.2 0.2 2 obedience 1.7 1.8 0.0 0.2 0.6 3.1 0.3 0.4 1 interesting job not imp. 0.0 1.3 19.1 0.1 0.0 0.3 17.4 3.0 2 interesting job 0.1 0.7 12.7 0.1 0.1 0.3 11.2 2.1 1 accept outgroups 0.2 1.0 1.9 0.8 0.0 1.7 1.0 0.9 2 reject outgroups 0.9 3.1 6.4 2.7 0.0 5.8 3.6 2.7 1 childless woman not OK 1.0 4.5 0.2 0.8 0.0 5.1 0.1 1.0 2 childless woman OK 1.0 6.9 0.2 1.3 0.0 7.1 0.1 1.4 1 single mother not OK 2.7 0.4 0.6 5.7 0.9 6.0 1.7 0.1 2 single mother OK 2.4 0.9 0.5 6.7 0.5 7.1 1.4 0.1 1 no thrift 0.0 3.5 0.3 0.0 0.2 1.3 0.2 1.8 2 thrift 0.0 7.0 0.5 0.0 0.4 2.8 0.4 3.3 1 no life after death 5.4 3.9 0.0 0.8 8.1 0.0 0.5 0.1 2 life after death 5.0 3.6 0.0 0.8 7.4 0.0 0.3 0.1 1 God not imp. 11.5 2.0 0.1 2.0 14.3 0.2 0.0 0.1 2 God imp. 5.9 1.0 0.1 1.0 7.3 0.1 0.0 0.0 1 no comfort from religion 11.3 1.9 0.1 1.5 13.8 0.4 0.0 0.0 2 comfort from religion 6.7 1.1 0.1 0.9 8.1 0.2 0.0 0.0 1 divorce not OK 3.7 1.2 0.0 4.4 1.4 7.5 0.1 0.2 2 divorce OK 3.3 1.8 0.0 4.8 1.0 8.1 0.0 0.1 1 achievement not imp. 0.1 3.0 2.4 5.0 0.0 0.0 0.1 10.0 2 achievement most imp. 0.5 7.9 6.1 13.3 0.1 0.2 0.3 26.5 1 work/life balance not imp. 0.3 0.3 21.1 3.7 0.2 0.0 7.8 17.5 2 work/life balance most imp. 0.1 0.2 11.7 2.1 0.0 0.0 4.2 9.7 1 good pay not imp. 0.0 3.6 4.4 7.0 0.0 0.0 12.4 1.7 2 good pay most imp. 0.0 6.2 7.3 12.1 0.0 0.1 21.0 2.9 1 fighting rising prices not most imp. 0.1 0.7 0.0 1.2 0.0 0.0 0.6 1.2 2 fighting rising prices most imp. 0.2 3.5 0.0 4.3 0.1 0.2 2.5 5.0 1 no subversive action 0.9 4.6 0.5 0.9 0.0 5.2 1.3 0.1 2 subversive action 1.2 5.4 0.6 1.2 0.1 6.4 1.5 0.1 1 service: never 6.2 0.8 0.2 6.7 8.9 0.1 0.6 1.1 2 service: occasional 0.7 0.5 0.0 6.5 0.3 1.5 1.4 3.3 3 service: 1+/month 7.6 1.8 0.1 0.2 8.2 1.0 0.3 0.7 Contribution exceeding the average as defined by dividing 100 by the total number of modalities (100/39 = 2.56%) are highlighted in bold. 31 Figure 5 Cloud of individuals from four societies in plane 1-2 (upper map) and 3-4 (lower map) with landmarks -2.5 -1.5 -0.5 0.5 1.5 2.5 Authoritarian (+) 2.5 122 955 239 119 1.5 Axis 2 [y] 0.5 -0.5 (-) Libertarian 189 046 -1.5 Russia '95 Jordan '01 Colombia '98 Sweden '99 104 551 -2.5 (-) Secular Axis 1 [x] Religious (+) World Values Survey, Russia (1995, N=1.000), Colombia (1998, N=1.000), Sweden (1999, N=1.013) and Jordan (2001, N=1.000). Data points represent individuals’ position along the four rotated axes of cultural values revealed by multiple correspondence analysis. The original samples for Russia, Colombia and Jordan include 2.038, 2.996 and 1.223 cases respectively: for ease of presentation, 1.000 cases were randomly selected preserving the regional breakdown of these original samples. Highlighted data points represent landmark individuals (marked using their unified respondent number in the WVS) whose description helps illustrate the four dimensions. -2.5 -1.5 -0.5 0.5 1.5 2.5 (-) Achievement not imp. Axis 4 [y] Achievement imp. (+) 3.0 104 551 2.0 1.0 189 046 0.0 -1.0 Russia '95 Jordan '01 Colombia '98 Sweden '99 122 955 239 119 -2.0 (-) Non-materialist Axis 3 [x] 32 Materialist (+) Figure 6 Per capita GDP v. Inglehart's instruments and scores from MCA Materialism score from rotated MCA solution 1.5 Materialism (Axis 4, rotated MCA solution) Inglehart's Postmaterialism Index* Inglehart's Postmaterialism Index (standardized reverse scale) 1.0 0.5 0.0 -0.5 R² = 0.4611 -1.0 -1.5 0 10 000 20 000 30 000 40 000 1.5 R² = 0.1933 1.0 0.5 0.0 -0.5 R² = 0.1016 -1.0 -1.5 0 10 000 2005 Real GDP/capita (PPP) 20 000 30 000 40 000 2005 Real GDP/capita (PPP) Last available data from the World Values Survey and per capita GDP figures (2005 US dollars at purchasing power parity) from the Penn World Table (Heston, Summers, and Aten 2002). The number of countries is 76 for the left and 81 for the right diagrams (excluding Luxembourg as influential case because of its high per capita GDP). Countries marked with round white data points have recently democratized ("post-authoritarian") governments. Authoritarianism Authoritarianism score from rotated MCA solution 1.5 1.0 0.5 0.0 1.5 -1.0 1.0 -1.5 R² = 0.4808 0 10 000 0.5 20 000 30 000 40 000 2005 Real GDP/capita (PPP) 0.0 Religiosity score from rotated MCA solution -0.5 -1.0 Religiosity Traditional (+) v. Secular/Rational (-) values Inglehart's Traditional-secular score -0.5 R² = 0.1774 -1.5 0 10 000 20 000 30 000 1.5 1.0 0.5 40 000 2005 Real GDP/capita (PPP) 0.0 -0.5 R² = 0.1526 -1.0 -1.5 0 10 000 20 000 30 000 40 000 2005 Real GDP/capita (PPP) 33 Appendix Table 5 Recoding of questions from the WVS for multiple correspondence analysis WVS name label (for composite scores, original variables are in italic) original scale in WVS a025 With which of these two statements do you tend to agree? A) One most always love and respect parents (regardless of their qualities and faults). B) Love and respect must be earned. 1: agree with A 2: agree with B 3: neither a040 Qualities that children can be encouraged to learn at home. Which, if any, do you consider to be especially important? a129 Which (if any) of these people would you not like to have as neighbors? a132 c011 always respect parents (2) respect parents conditional (1) imagination (0, DK) imagination not imp. (1) thrift (0, DK) no thrift (1) learn faith imp. (0, DK) learn faith not imp. (1) obedience (0, DK) no obedience (0) accepts outgroups (1, 2, 3) rejects outgroups (1) good pay imp. (0) good pay not imp. (0) work/life balance not imp. (1, 2, 3) work/life balance imp. (0) achievement not imp. (1 thru 4) achievement imp. Thrift, saving money and things 1: important 0: not mentioned Religious faith Immigrants/foreign workers People who have AIDS 0 to 3, depending on the number of groups rejected Homosexuals 1: mentioned 0: not mentioned Important in a job: good pay c012 c015 (1) Obedience a042 a130 label for modalities in MCA Imagination a034 a038 recoding rule (DK=Don't know) Not too much pressure Important in a job: Good hours c017 Generous holidays c014 Respected job c016 Opportunity to use initiative Important in a job: c018 You can achieve something c019 Responsible job 0 to 3, depending on the number of items mentioned 0 to 4, depending on the number of items mentioned 34 WVS name label (for composite scores, original variables are in italic) c020 Important in a job: a job that's interesting? d019 d023 e003 e026 e028 e029 f028 f051 f063 f064 f121 recoding rule (DK=Don't know) label for modalities in MCA (1) interesting job (0) interesting job. not imp. (1) childless woman not OK (2) childless woman OK 0: disapprove 1: approve 2: depends (0) single mother not OK (1) single mother OK 1: maintaining order in the nation 2: give people more say 3: fighting rising prices 4: protecting freedom of speech (3) fight rising prices most imp. (1, 2, 3, DK) fight rising prices not most imp. (if 3 or 2 for any) subversive action (if no 3 or 2 for any: 0, DK) no subversive action (1, 2, 3) service: 1+/month (4, 5, 6, 7) service: occasional (8) service: never (1) life after death (0, DK) no life after death (1 thru 5, DK) God not imp. (6 thru 10) God imp. (1) comfort from religion (0, DK) no comfort from religion not OK (1 thru 4) divorce not OK OK (6 thru 10) divorce OK original scale in WVS 1: mentioned 0: not mentioned 1: needs children 2: not necessary Woman has to have children in order to be fulfilled? Woman wants to have child as a single parent but doesn't want to have a stable relationship with a man If you had to choose, which one of the things would you say is most important? Of different forms of political action, whether you have actually done, might do or would never do any of these things: Joining in boycotts 1: have done 2: might do 3: would never do Joining unofficial strikes Occupying buildings or factories 1: 2+/week 2: weekly 3: monthly 4: special holydays 5: other specific holy days 6: once a year 7: less often 8: never, practically never How often do you attend religious services? 1: yes 0: no Do you believe in live after death? How important is God in your life? 1 (not at all) to 10 (very) 1: yes 0: no Do you get comfort and strength from religion? 1 (never justifiable) to 10 (always justifiable) Divorce All variables are recoded into binary modalities except f028 (attendance of religious services), recoded into three categories. "Don't know" answers are used for a number of questions, as specified in the recoding rules. Variables c011 to c019 (work attitudes) are combined into new variables based on results from a separate multiple correspondence analysis revealing a twodimensional latent structure. 35 References Adachi, Kohei. 2004. “Oblique Promax rotation applied to the solutions in multiple correspondence analysis.” Behaviormetrika 31:1–12. Bennett, Tony, Mauricio Bustamante, and John Frow. 2013. “The Australian space of lifestyles in comparative perspective.” Journal of Sociology 49:224–255. Blasius, Jörg, and Peter Graeff. 2009. “Economic freedom, wealth and corruption: Analyzing their empirical connection by means of correspondence analysis.” Pp. 135–155 in Graeff, Peter, and Mehlkop, Guido: Capitalism, democracy and the prevention of war and poverty. Blasius, Jörg, and Michael Greenacre. 2006. “Correspondence Analysis and Related Methods in Practice.” Pp. 3–40 in Blasius, Jörg, and Greenacre, Michael: Multiple Correspondence Analysis and Related Methods. Boca Raton, FL: Chapman & Hall/CRC. Blasius, Jörg, and Andreas Mühlichen. 2010. “Identifying audience segments applying the ‘social space’ approach.” Poetics 38:69–89. Blasius, Jörg, and Victor Thiessen. 2006a. “A three-step approach to assessing the behaviour of survey items in cross-national research.” Pp. 433–453 in Blasius, Jörg, and Greenacre, Michael: Multiple Correspondence Analysis and Related Methods. Boca Raton, FL: Chapman & Hall/CRC. Blasius, Jörg, and Victor Thiessen. 2006b. “Assessing Data Quality and Construct Comparabilityin Cross-National Surveys.” European Sociological Review 22:229–242. Bourdieu, Pierre. 1987. Choses dites. Paris: Les Éditions de Minuit. 36 Bourdieu, Pierre. 1979. La distinction. Critique sociale du jugement. Paris: Les Éditions de Minuit. Bourdieu, Pierre. 2000. Les structures sociales de l’économie. Paris: Seuil. Bourdieu, Pierre. 1999. “Une révolution conservatrice dans l’édition.” Actes de la recherche en sciences sociales 126-127:3–28. Bourdieu, Pierre, and Monique De Saint Martin. 1978. “Le patronat.” Actes de la recherche en sciences sociales 20-21:3–82. Böltken, Ferdinand, and Wolfgang Jagodzinski. 1985. “In an Environment of Insecurity: Postmaterialism in the European Community, 1970 to 1980.” Comparative Political Studies 17:453–484. Buhlmann, Felix, Thomas David, and André Mach. 2013. “Cosmopolitan Capital and the Internationalization of the Field of Business Elites: Evidence from the Swiss Case.” Cultural Sociology 7:211–229. Chalcraft, David J. 2005. “Reading Weber’s Patterns of Response to Critics of The Protestant Ethic: Some ‘Affinities’ in and between Replies to Felix Rachfahl and Werner Sombart.” Journal of Classical Sociology 5:31–51. Chalcraft, David J., and Austin (eds.) Harrington. 2001. The Protestant ethic debate: Max Weber’s replies to his critics, 1907-1910. Liverpool: Liverpool University Press. Clarke, Harold D., and Nitish Dutt. 1991. “Measuring Value Change in Western Industrialized Societies: The Impact of Unemployment.” The American Political Science Review 85:905. Clarke, Harold D., Nitish Dutt, and Jonathan Rapkin. 1997a. “(Mis)Measuring Value Change: A Brief Postscript.” Political Behavior 19:61–64. 37 Clarke, Harold D., Nitish Dutt, and Jonathan Rapkin. 1997b. “Conversations in context: The (mis) measurement of value change in advanced industrial societies.” Political Behavior 19:19–40. Davidov, Eldad, Peter Schmidt, and Shalom H. Schwartz. 2008. “Bringing Values Back In: The Adequacy of the European Social Survey to Measure Values in 20 Countries.” Public Opinion Quarterly 72:420–445. Davis, Darren W., and Christian Davenport. 1999. “Assessing the Validity of the Postmaterialism Index.” Political Science 93:649–664. Davis, Darren W., Kathleen M. Dowley, and Brian D. Silver. 1999. “Postmaterialism in World Societies: Is It Really a Value Dimension?” American Journal of Political Science 43:935 – 962. Deth, Jan W. Van. 1983. “The Persistence of Materialist and Post-Materialist Value Orientations.” European Journal of Political Research 11:63–79. Duch, Raymond M., and Michaell A. Taylor. 1994. “A Reply to Abramson and Inglehart’s ‘Education, Security, and Postmaterialism.’” American Journal of Political Science 38:815. Duch, Raymond M., and Michaell A. Taylor. 1993. “Postmaterialism and the Economic Condition.” American Journal of Political Science 37:747. Flanagan, Scott C. 1982a. “Changing Values in Advanced Industrial Societies: Inglehart’s Silent Revolution from the Perspective of Japanese Findings.” Comparative Political Studies 14:403–444. Flanagan, Scott C. 1982b. “Measuring Value Change in Advanced Industrial Societies: A Rejoinder to Inglehart.” Comparative Political Studies 15:99–128. Flemmen, Magne. 2012. “The Structure of the Upper Class: A Social Space Approach.” Sociology 46:1039–1058. 38 Graaf, Nan Dirk De, and Geoffrey Evans. 1996. “Why are the Young more Postmaterialist?: A Cross-National Analysis of Individual and Contextual Influences on Postmaterial Values.” Comparative Political Studies 28:608–635. Granato, Jim, Ronald Inglehart, and David Leblang. 1996a. “Cultural Values, Stable Democracy, and Economic Development: A Reply.” American Journal of Political Science 40:680–696. Granato, Jim, Ronald Inglehart, and David Leblang. 1996b. “The Effect of Cultural Values on Economic Development: Theory, Hypotheses, and Some Empirical Tests.” American Journal of Political Science 40:607–631. Haller, Max. 2002. “Theory and Method in the Comparative Study of Values: Critique and Alternative to Inglehart.” European Sociological Review 18:139–158. Hanquinet, Laurie. 2013. “Visitors to modern and contemporary art museums: towards a new sociology of ‘cultural profiles.’” The Sociological Review 61:790–813. Hanquinet, Laurie, Henk Roose, and Mike Savage. 2013. “The Eyes of the Beholder: Aesthetic Preferences and the Remaking of Cultural Capital.” Sociology 48:111–132. Harrits, Gitte Sommer, Annick Prieur, Lennart Rosenlund, and Jakob Skjott-Larsen. 2010. “Class and Politics in Denmark: Are Both Old and New Politics Structured by Class?” Scandinavian Political Studies 33:1–27. Hendricks, J., and C. Breckinridge Peters. 1973. “The Ideal Type and Sociological Theory.” Acta Sociologica 16:31–40. Heston, Alan, Robert Summers, and Bettina Aten. 2002. “Penn World Table.” Center for International Comparisons at the University of Pennsylvania. Hjellbrekke, Johs. et al. 2007. “The Norwegian Field of Power Anno 2000.” European Societies 9:245–273. 39 Howe, Richard Herbert. 1978. “Max Weber’s Elective Affinities: Sociology Within the Bounds of Pure Reason.” The American Journal of Sociology 84:366–385. Inglehart, Ronald. 1997. Modernization and Postmodernization. Cultural, Economic and Political Change in 43 Societies. Princeton, New Jersey: Princeton University Press. Inglehart, Ronald. 1983. “The Persistence of Materialist and Post-Materialist Value Orientations: Comments on Van Deth’s Analysis.” European Journal of Political Research 11:81–91. Inglehart, Ronald. 1971. “The Silent Revolution in Europe: Intergenerational Change in PostIndustrial Societies.” The American Political Science Review 65:991–1017. Inglehart, Ronald. 1977. The Silent Revolution: Changing Values and Political Styles among Western Publics. Princeton University Press. Inglehart, Ronald, and Paul R. Abramson. 1994. “Economic Security and Value Change.” The American Political Science Review 88:336–354. Inglehart, Ronald, and Wayne E. Baker. 2000. “Modernization, cultural change, and the persistence of traditional values.” American Sociological Review 65:19–51. Inglehart, Ronald, and Scott C. Flanagan. 1987. “Value change in industrial societies.” The American Political Science Review 81:1289–1319. Kaiser, Henry F. 1958. “The varimax criterion for analytic rotation in factor analysis.” Psychometrika 23:187–200. Kiers, Henk A. L. 1991. “Simple structure in component analysis techniques for mixtures of qualitative and quantitative variables.” Psychometrika 56:197–212. Kropp, Kristoffer. 2013. “Social sciences in the field of power - the case of Danish social science.” Social Science Information 52:425–449. 40 Lafferty, William M., and Oddbjørn Knutsen. 1985. “Postmaterialism in a Social Democratic State: An Analysis of the Distinctness and Congruity of the Inglehart Value Syndrome in Norway.” Comparative Political Studies 17:411–430. Lebaron, Frédéric, and Brigitte Le Roux. 2011. “Social capital in the field of power : the case of Norway.” Review Literature And Arts Of The Americas 1:86–108. Lorenzo-Seva, Urbano, Michel van de Velden, and Henk A. L. Kiers. 2009. “CAR: A MATLAB Package to Compute Correspondence Analysis with Rotations.” Journal of Statistical Software 31:1–14. Löwy, Michael. 2004. “Le concept d’affinité élective chez Max Weber.” Archives de sciences sociales des religions 93–103. Lucas, Robert E. 1988. “On The Mechanics of Economic Development.” Journal of Monetary Economics 22:3–42. Majima, Shinolu, and Mike Savage. 2007. “Have There Been Culture Shifts in Britain?: A Critical Encounter with Ronald Inglehart.” Cultural Sociology 1:293–315. Marsh, Alan. 1975. “The ‘Silent Revolution,’ Value Priorities, and the Quality of Life in Britain.” The American Political Science Review 69:21–30. Martin, John Levi. 2003. “What Is Field Theory?” American Journal of Sociology 109:1–49. Maslow, Abraham Harold. 1970. Motivation and Personality. New York: Harper & Row. McKinnon, Andrew M. 2010. “Elective Affinities of the Protestant Ethic: Weber and the Chemistry of Capitalism.” Sociological Theory 28:108–126. Mills, Charles Wright. 1940. “Situated Actions and Vocabularies of Motive.” American Sociological Review 5:904–913. Purhonen, Semi, and David Wright. 2013. “Methodological Issues in National-Comparative Research on Cultural Tastes: The Case of Cultural Capital in the UK and Finland.” Cultural Sociology 7:257–273. 41 Romer, Paul M. 1990. “Endogenous Technological Change.” The Journal of Political Economy 98:S71–S102. Rouanet, Henry, Wemer Ackermann, and Brigitte Le Roux. 2000. “The Geometric Analysis of Questionnaires: the Lesson of Bourdieu’s La Distinction.” Bulletin de Méthodologie Sociologique 65:5–18. Roux, Brigitte Le, Henry Rouanet, Mike Savage, and Alan Warde. 2008. “Class and Cultural Division in the UK.” Sociology 42:1049–1071. Runciman, W. G. 2005. “Not Elective but Selective Affinities.” Journal of Classical Sociology 5:175–187. Saint Martin, Monique De, and Pierre Bourdieu. 1987. “Agrégation et ségrégation [Le champ des grandes écoles et le champ du pouvoir].” Actes de la recherche en sciences sociales 69:2–50. Saracco, Jérôme, Marie Chavent, and Vanessa Kuentz. 2010. “Rotation in Multiple Correspondence Analysis: a planar rotation iterative procedure.” Working Papers pf GREThA 04. Savage, Mike, and Elizabeth B. Silva. 2013. “Field Analysis in Cultural Sociology.” Cultural Sociology 7:111–126. Schwartz, Shalom H. 2011. “Studying Values: Personal Adventure, Future Directions.” Journal of Cross-Cultural Psychology 42:307–319. Silva, Elizabeth B., and Brigitte Le Roux. 2011. “Cultural capital of couples: Tensions of elective affinities.” Poetics 39:547–565. Silva, Elizabeth B., and David Wright. 2009. “Displaying Desire and Distinction in Housing.” Cultural Sociology 3:31–50. Solow, Robert M. 1956. “A Contribution to the Theory of Economic Growth.” The Quarterly Journal of Economics 70:65–94. 42 Swan, Trevor W. 1956. “Economic Growth and Capital Accumulation.” Economic Record 32:334–361. Treviño, A. Javier. 2005. “Parsons’s Action-System Requisite Model and Weber's Elective Affinity: A Convergence of Convenience.” Journal of Classical Sociology 5:319–348. Trump, Thomas M. 1991. “Value Formation and Postmaterialism: Inglehart’s Theory of Value Change Reconsidered.” Comparative Political Studies 24:365–390. Veenstra, Gerry. 2007. “Social space, social class and Bourdieu: health inequalities in British Columbia, Canada.” Health & place 13:14–31. Velden, Michel, and Henk A. L. Kiers. 2005. “Rotation in Correspondence Analysis.” Journal of Classification 22:251–271. Vijver, F. J. R. van De, and Ype H. Poortinga. 2002. “Structural Equivalence in Multilevel Research.” Journal of Cross-Cultural Psychology 33:141–156. Wacquant, Loïc. 2013. “Symbolic power and group-making: On Pierre Bourdieu’s reframing of class.” Journal of Classical Sociology 13:274–291. Weber, Max, Peter R. Baehr, and Gordon C. Wells. 2002. The Protestant Ethic and the “Spirit” of Capitalism and Other Writings. New York: Penguin Classics. Welzel, Christian. 2002. “Effective Democracy, Mass Culture, and the Quality of Elites: The Human Development Perspective.” International Journal of Comparative Sociology 43:317–349. Welzel, Christian, Ronald Inglehart, and Hans-Dieter Kligemann. 2003. “The theory of human development: A cross-cultural analysis.” European Journal of Political Research 42:341–379. World Values Survey Association. 2009. “1981-2008 Official Aggregate v.20090901 Aggregate File Producer: ASEP/JDS, Madrid (www.worldvaluessurvey.org).” 43