POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? DIMENSIONALITY IN INGLEHART'S MATERIALISM- POSTMATERIALISM SCALE*

Transcription

1 International Journal of Public Opinion Research Vol. 6 No /94 S3.00 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? DIMENSIONALITY IN INGLEHART'S MATERIALISM- POSTMATERIALISM SCALE* Clive Bean and Elim Papadakis ABSTRACT Ronald Inglehart's Postmaterialism thesis has undoubtedly been one of the most influential ideas in cross-national and cross-temporal research on political behavior over the past two decades. The wide-ranging debate and criticism generated by the Inglehart thesis has focused both on theoretical issues concerning the nature of values and on methodological issues concerning the measurement of Materialism and Postmaterialism, particularly the method for ranking individual responses. Using a data set that employed both the original ranking method for measuring values and an alternative rating method, we explore the dimensional structure of these values and discuss the implications of differences between the two measurement strategies. The two-dimensional solution provided by the rating method is, we argue, a more theoretically appropriate way of understanding Materialist and Postmaterialist values than the notion of a single conflict dimension, since it allows for a more flexible and realistic account of the choices made by most social actors choices which may represent both Materialism and Postmaterialism. The enduring interest in the work of Ronald Inglehart is founded on his claims about the emergence of new socio-political cleavages in industrialized societies. The rise of social movements (like those concerned about environment protection) and their influence on established institutions appears to confirm many of the presuppositions of Inglehart about a 'silent revolution' in values among western publics. The preoccupation with values has been central to the work of Inglehart and of writers on social movements and the 'new polities'. They have posited a decisive shift from a concern about economic growth and material prosperity to a focus on spiritual and non-material values. Like Inglehart, * Revised version of a paper presented to the Australasian Political Studies Association conference, Canberra, September-October 1992 and to the International Sociological Association's Research Committee on Social Stratification and Mobility (RC 28) meeting, Canberra, January We thank participants for valuable comments. & World Association for Public Opinion Research igq4

2 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 265 writers who have emphasized the novel qualities of social movements have posited a fundamental conflict between proponents of economic and non-economic values (see Cotgrove and Duff, 1981). For instance, the proponents of economic values may experience great difficulty in adapting to the demands of those who emphasi2e environmental values, including ideas about the intrinsic value of nature or about a fundamental tension between material and spiritual values. Our perspective is that it is not necessary to accept all the claims about a fundamental conflict between Material and Postmaterial values. Social movements and other organizations may draw on both a dominant way of thinking oriented towards economic growth and alternative approaches. Furthermore, the causal linkages between values and social change or social conflict will vary from context to context. Though mainly concerned with values, Inglehart (1977) has at least been aware of the economic, technological developments that led to the satisfaction of needs for sustenance by a large proportion of the population and of social changes that may have influenced people's orientations towards the future (including rising levels of education, distinctive experiences of different generations, the expansion of mass communications and the increase in geographic mobility). Writers who emphasize values are also aware that participation in new social movements may depend on the existence of objective problems (like environmental degradation) and of appropriate organizational structures (like social networks and voluntary associations). However, rather than analyzing the precise connection between values and economic and social changes, writers like Inglehart have concentrated on examining the significance of values and preferences (for the development of social movements, for instance). The rationale for studying values and preferences is that they are preconditions for social action. Inglehart has argued that 'people do not act unless they want to attain some goal' and that the 'existence of problems and organizations would have no effect unless some value system or ideology motivated them to act' (1990, p. 371). The way people perceive issues is of course bound to have a decisive impact on whether or not they feature on the political agenda. Inglehart has argued that changes in values and political skills have had major consequences for politics and society. They have contributed to changes in the political agenda (like a greater concern for 'life-style' and quality of life issues), to a decline in the significance of class conflict and to a change in the social bases of political conflict, to changes in support for established national institutions and to a decline in the legitimacy of the nation state. They have also apparently influenced the style of political participation, leading to new challenges to elites by movements that emphasize participation and specific policy issues.

3 266 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH THE MEANING AND SIGNIFICANCE OF INGLEHART'S WORK Initially, Inglehart (1971) measured Materialist and Postmaterialist values by presenting survey respondents with a list of four goals and then asking them to choose the two most desirable. The four items were 'maintaining order in the nation', 'giving the people more say in important political decisions', 'fighting rising prices' and 'protecting freedom of speech'. The second and fourth items were designed to find out who identified with the protests in the 1960s for more civil rights and liberty. Respondents who chose these two items were classed as Postmaterialist (or 'post-bourgeois', in Inglehart's original terminology). Those who chose 'maintaining order' and 'fighting rising prices' were deemed to be 'acquisitive', or Materialist. A third 'mixed' category was used for those (the majority) who were neither outright Materialists nor Postmaterialists. Inglehart has since added eight more items to the list, as discussed below. The method of presentation, including the introduction and exact wording of the items, has varied in subsequent surveys. The central argument is that Postmaterialists are more likely than Materialists to take economic security for granted and to emphasize aesthetic and intellectual development rather than the pursuit of material goals. The scale has been useful in identifying the tendency by some people to stress participation and individual freedom rather than economic and personal security. It has also been used to assess the association between values and political behavior as well as economic and social circumstances (see Inglehart, 1990). For example, there is an apparent connection between Postmaterialist values and support for political parties like the Australian Democrats who, among the established parties in Australia, have had the most pro-environmentalist platform; similarly, supporters and sympathizers of groups campaigning for environmental protection are much more likely to be Postmaterialists than those who are either indifferent towards or opposed to such groups; and finally, there is also a strong link between levels of education, Postmaterialist values and support for environmentalism (see Papadakis, 1993). Inglehart initially attempted to justify the use of the scale by referring to sociological and psychological theories about needs and preferences. The notions of scarcity and socialization are used to support the distinction between Materialism and Postmaterialism. The idea of scarcity is derived from the 'hierarchy of needs' postulated by the psychologist Maslow (1954). Primary or basic needs for survival (like food, shelter, and security) are located at the lower end of this hierarchy. They are labeled material needs.

4 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 267 Once they have been met, people are more likely to seek fulfillment of secondary needs, including intellectual, aesthetic, and social needs. The hierarchy of needs can be further broken down into four subgroups (Inglehart, 1990, p. 134), two Materialist (economic security and physical security) and two Postmaterialist (belonging and esteem, and aesthetic and intellectual needs). In affluent societies basic material needs for food and shelter are usually met, if not by individuals receiving income as wage earners, at least by the welfare state, which is supposed to provide a basic safety net. People in affluent societies are therefore more likely than those in societies experiencing relatively high levels of material deprivation to be able to pursue intellectual, aesthetic, and social goals of a Postmaterialist nature. However, during economic recessions, there is likely to be a diminishing emphasis on Postmaterial values. Inglehart has postulated that this decline is not, however, inevitable. The effect of scarcity on values can be held in check by the socialization of individuals into certain values during particular phases of the life cycle. In other words, Inglehart has presented a parallel hypothesis based on the notion of socialization. The argument is simply that the nature of people's upbringing (or, socialization) prior to adulthood will have a lasting impact and will be relatively independent of changes in economic fortunes and political upheavals. People who are socialized in different historical eras (which may be characterized either by affluence or deprivation or economic decline) will later in life each react differently to the economic, social, and political environment. Studies of political behavior have shown, for example, that many people have continued to identify with the same political party from their youngest years through to old age (for example, Campbell et al., i960; Butler and Stokes, 1974). Several studies have shown significant differences in value priorities between people who grew up during the Great Depression and those who were raised during the affluent post-war years (for instance, Inglehart, 1981; Baker et ai, 1981). Those who experienced the Great Depression were more likely than younger age cohorts to have Materialist values, even following improvements in socioeconomic conditions and radical changes in the political system. Though Inglehart was initially concerned about the protest movements of the 1960s (the student and civil rights movements), numerous studies have shown that many of the supporters of these movements played leading roles in the social movements of the 1970s. The earlier movements had also anticipated some of the preoccupations of the green movements in the 1970s and 1980s. The analysis by Inglehart of the impact of these earlier movements has thus proved of lasting value.

5 268 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH THE QUESTION OF DIMENSIONALITY: CONCEPTUAL ISSUES Our principal concern is with the suggestion by Inglehart that Materialism and Postmaterialism represent fundamentally different orientations, in other words that they represent a new basis for conflict. Using a variety of statistical evidence, Inglehart has argued that Materialism and Postmaterialism are mutually exclusive, in other words that there are distinct and consistent clusters in people's preferences or values. For instance, by performing a principal components analysis on respondents' rankings of the full twelve-item battery of Materialist and Postmaterialist goals in ten countries, Inglehart found that the six Materialist items tended to be chosen together and that the Postmaterialist items (with one exception) were also grouped together, opposite the Materalist items (Inglehart, 1977, p. 44). Using the same technique, he also finds that these groupings hold over time (Inglehart, 1979, p. 316; 1990, p. 139). In response to his critics, who have argued for much greater differentiation between the items used in the scale, Inglehart has insisted that his twelve items unequivocally 'tap an underlying Materialist/Postmaterialist dimension' (Inglehart, 1990, p. 144). Prior to reanalyzing these items, we will address a number of conceptual issues. First, there is nothing essentially novel about the contrast between Material and Postmaterial values. Aristotle had suggested that once people had satisfied their basic material needs they might aspire to fulfill 'higher' spiritual needs. Similarly, in the short history of socialism, the revisionists since Bernstein (1899) have made similar claims. Second, the Materialism-Postmaterialism scale oversimplifies the complexity of individual preferences and values. The scale fails to consider areas of social life that lie outside the political and the economic, for example, values related to religion and the family (Herz, 1979). It also omits important items from within the political and economic realms, such as unemployment (Clarke and Dutt, 1991). Inglehart (1990, p. 133) has himself conceded that Materialism-Postmaterialism is not the only dimension along which cultural change can take place and others have demonstrated how the Materialist-Postmaterialist dimension may fit into broader conceptions of cultural orientation (Hellevik, 1993). Third, some have questioned whether or not his scale represents values. Muller-Rommel (1983, p. 223), for instance, has suggested that the scale may represent attitudes and that Inglehart has confused the two. Nonetheless, there does appear to be a case for arguing that Inglehart has used the term values correctly, particularly if one defines values as 'the most important beliefs' a person holds and beliefs as 'anything the individual thinks is, might be or ought to be' (see Hennessy, 1985, PP- 4-5)-

6 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 269 Many of these criticisms are directed at the inappropriateness of the Postmaterialist thesis for understanding factors that actually shape social change, notably the institutions that either constrain the scope for rapid social change or steer it in certain directions. Though Inglehart (1990) has successfully challenged some of the findings that have attempted to disprove his claims about a rise in Postmaterialism, there hangs a question mark over his central assumption of a basic conflict between Materialism and Postmaterialism. We accept that the method used by Inglehart is valuable in identifying factors that may have influenced social change. However, we also wish to draw attention to some of the difficulties with assuming that there is a single Materialist- Postmaterialist conflict dimension. Faced with sustained criticism of his method of ranking values (for example, by Herz, 1979 and Flanagan, 1987), Inglehart has responded by citing studies which have shown that the division between Materialists and Postmaterialists has remained an important characteristic of varying perceptions of the world among citizens in western democracies (for example, Marsh, 1977; Lafferty and Knutsen, 1985, cited in Inglehart, 1990, pp ). He has also dismissed as anomalies the findings by Lafferty (1975), Van Deth (1983a) and Boltken and Jagodzinski (1985) about a decline in Postmaterialism. His own findings, with much larger samples, show a consistent difference between younger and older birth cohorts in their value orientations (Inglehart, 1990, pp. 90-1). The main explanation for any decline in Postmaterialism at a particular period is the growth of inflation. Though he has acknowledged period effects, Inglehart has stood by his claim about the rise in Postmaterialism and demonstrated how, between 1970 and 1988, the ratio of Materialists to Postmaterialists changed from 4:1 to 4:3 among the publics of six west European nations (Inglehart, 1990, pp ). In the face of this evidence it would be unwise to deny that Inglehart's scale provides a useful basis for understanding some of the most interesting changes in political behavior. However, we are concerned that, in defending himself against his critics, Inglehart has overlooked some basic issues concerning the relationship between Materialism and Postmaterialism and has too hastily posited a single conflict dimension between two sets of values. The argument for a single dimension even contradicts one of the pillars on which Inglehart's theory is built, namely the notion of scarcity and the 'hierarchy of needs'. The hierarchy suggests that people become Postmaterialists only after basic material needs have been met. Postmaterialists are therefore, logically, according to this argument, also Materialists. There is no necessary fundamental conflict between these orientations. Interestingly, several writers (Marsh, 1977; Cotgrove, 1982, p. 47; Milbrath, 1984, pp. 24,28) have recognized that support for Postmaterialist goals does not necessarily imply an abandonment of Materialist ones.

7 270 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH Inglehart has not failed to recognize that many people espouse Materialist and Postmaterialist goals (1979, p. 320). Most people are neither outright Materialists nor Postmaterialists. Inglehart has also stressed that Material security takes precedence over Postmaterialism (1979, p. 309). However, he has not taken up the challenge of explaining the apparent inconsistency between claims about a single conflict dimension and the argument about how support for Postmaterialism does not necessarily imply the abandonment of Materialist goals. This challenge has been taken up by writers like Cotgrove and Duff (1981) who have argued that different values are relevant to different contexts and suggested that notions like Materialism and Postmaterialism serve as a heuristic device rather than a definitive guide to the source of conflict in society. Again, Inglehart implicitly recognizes the problem with his claim about a single conflict dimension when he argues that Postmaterialism can be self-defeating, in other words that the 'antiindustrial outlook of some of the movement's ideologues could lead to neglect of the economic base on which Postmaterialism ultimately depends' (1990, p. 334). Inglehart predicts the emergence of a 'new synthesis' of Materialist and Postmaterialist orientations 'through sheer functional necessity'. This possibility has recently been illustrated by the international environmental group, Greenpeace, which has in the past campaigned on the basis of assumptions about a fundamental conflict between environmental protection and economic development. In September 1992 Greenpeace's new international director, Paul Gilding, announced that the organization would switch its tactics from confrontation to cooperation with government and industry. It appears, after all, that the basic, single conflict dimension may not be so fundamental. THE QUESTION OF DIMENSIONALITY: EMPIRICAL EVIDENCE At one level, there is strong prima facie evidence in the data produced by Inglehart's four-item battery that the notion of a single polarized dimension does not reflect the value priorities of most citizens adequately. When asked to choose the two most important priorities of the four, only a minority of respondents in most western countries reveal themselves to be 'pure' Materialists or 'pure' Postmaterialists. In general, a substantial majority usually choose one Materialist priority and one Postmaterialist priority: in other words they emphasize both Materialist and Postmaterialist values together. In this respect, Australia, the source of the data for the current investigation, exhibits a very similar pattern to those found in other comparable countries (see Dalton, 1988, pp. 84-5; Inglehart, 1990, p. 93; Papadakis, 1991, p. 247). For example, in the 1988 national survey which forms the basis for this analysis, 27 percent of

8 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 271 Australians could be classed as Materialist, 13 percent Postmaterialist and 61 percent mixed. Until now, few researchers have addressed the issue of dimensionality in Materialist-Postmaterialist value orientations by directly comparing the utility and analytic implications of alternative ways of measuring the scale. Some scholars have obtained multi-dimensional solutions using methods which require respondents to rate each individual item in the battery separately (for example, Flanagan, 1987; Reimer, 1988), rather than using Inglehart's traditional ranking method. But in such cases they have not usually had ranking-based measures in the same data set with which to make direct comparisons and indeed the specific items do not always precisely replicate those of Inglehart and nor have the data always been based on mass population samples. These three qualifications apply also to Dalton's (1981) study which identified a single dimension from rating-based measures of values among young Germans. One partial exception is the work of Van Deth (1983^), which focused on a national Dutch sample containing both rankings and ratings of the twelve-item Inglehart battery. After extensive analysis of both the ranked and rated measures, involving a variety of statistical techniques, Van Deth concluded unequivocally that for both the ratings and rankings there are 'distinct materialist and post-materialist clusters... that are... not reduceable to a single dimension' (1983^, p. 424). It is somewhat mystifying to find Inglehart's interpretation of Van Deth's argument as being that 'he [Van Deth] concludes that the underlying dimension is quite robust, for he obtains a clear Materialist/Postmaterialist dimension from both rankings and ratings' (1990, p. 144). Van Deth did find the concepts of Materialism and Postmaterialism to be empirically robust, but as two separate dimensions, not as a single dimension (a point he emphasized repeatedly in the course of his article). Being mainly concerned with methodology, Van Deth did not proceed to investigate the substantive consequences of the discovery of two separate dimensions. We are fortunate to have a data set which includes both the traditional ranking method of measuring Materialism-Postmaterialism and an alternative rating method, as well as an extensive range of social structural variables and various measures of political behavior with which to test the performance of the differently constructed values scales. Having all these measures in the same data set has the dual advantage of permitting direct comparisons of the alternative measures of values (rather than the indirect comparisons across different surveys, and frequently across different countries, to which researchers are usually forced to resort) plus eliminating the inevitable doubts that normally linger in such exercises over the comparability of background variables in different data sources. Our data source is a national survey of Australians' attitudes to social welfare

9 272 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH and related issues, conducted in The survey was based on a systematic random sample drawn from the Australian national electoral rolls. Questionnaires were mailed out to prospective respondents and, after three follow-up mailings, the last of which was by certified mail, the final sample size achieved was 1,814, representing a response rate of 62 percent. 1 Supplementary data from the Australian National Social Science Survey (NSSS) are used to reinforce some of the arguments made on the basis of our core findings. 2 In the 1988 Australian survey the standard four-item values battery was presented to respondents first, followed by the additional eight items. This format follows one of the ways in which Inglehart has presented the items on different occasions (see Inglehart, 1979, p. 312). Respondents were initially asked to rate the items on a five-point scale labeled 'extremely important', 'very important', 'fairly important', 'not important', 'not at all important'. Then respondents were asked to choose among the items (thus producing rankings in the standard Inglehart manner), separately for the four- and eight-item batteries. The page from the questionnaire showing the exact format and wording of the questions is reproduced in the appendix. The NSSS used a very similar introduction, but listed all twelve items together (and in a slightly different order from the 1988 welfare survey) and only asked respondents to rate the items, not to rank them as well. The seven-point scale was labeled: 'the most important single thing', 'extremely important', 'very important indeed', 'fairly important', 'not important', 'undesirable, a bad idea', 'a very bad idea'. We proceed by employing principal components factor analysis to investigate the structure of the data, measured according to both the ranking and rating procedures. On the basis of the results we then construct appropriate scales in order to test their comparative analysis performance. While acknowledging that factor analysis is not strictly appropriate for data with ipsative properties, like the ranked items (Jackson and Alwin, 1980; Van Deth, 1983^), our concern is to replicate Inglehart's methodology as faithfully as possible when handling the ranked data, in order to address his work directly and to avoid any danger of reaching different results simply because we used different ' Initially, 3,507 questionnaires were mailed out. Non-contacts reduced the effective sample size by 564, leaving a sample of 2,043 fr m which the 1,814 completed replies were received. The survey was funded by the Social Policy Research Centre, University of New South Wales and the University of New England. It is publicly available through the Social Science Data Archives (SSDA) at the Australian National University (ANU). For further details see Papadakis (1090a). 2 The NSSS was conducted by Jonathan Kelley, Robert G. Cushing, and Bruce Headey (1987), funded by the Australian Research Grants Scheme and the Research School of Social Sciences, ANU, and distributed through the SSDA. It is a national survey of the Australian adult population (JV=3,0I2), conducted by personal interview in urban areas and by mail in rural areas. The response rate was just under 60 percent. Neither the original collectors of the data, nor the funding agencies, are responsible for the analyses or interpretations in this paper.

10 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 273 methods. Inglehart himself points out (1977, p. 43) that the problems of ipsivity are relatively minor when as many as twelve items are involved. Panel A of Table 1 shows the results of a factor analysis based on the standard twelve-item Inglehart battery and the Inglehart ranking method of measuring the variables. Scoring of the individual items is based on the descriptions in Inglehart (1979, PP315, 319), with items being scored according to whether they were chosen as the first or second most important aim among the initial four items, or according to whether they were chosen as first, second, or third most important among the second group of eight items. As one can see from the initial column of Table 1 (Panel A), which contains the first unrotated factor (the evidence Inglehart always presents in support of the unidimensionality of the Materialism-Postmaterialism scale), at this early stage the data reassuringly produce a similar pattern to that which Inglehart (1977; 1979; 1990) himself repeatedly gets from at least ten different European countries and the United States of America. The six Materialist items form a cluster with negative loadings while the six Postmaterialist items form a cluster with positive loadings (and unlike most of Inglehart's own results, the 'make our cities and countryside more beautiful' item does not stand out as having a much weaker loading than all of the other items although, like several of the other items, its loading is certainly not high). This factor also explains a similar amount of the variance to the first factor as reported in Inglehart's (1977, p. 44) own analyses (18 percent). Inglehart admits that the items tend to be spread over several dimensions when the matrix is rotated (1977, p. 43; 1979, PP ), although he does not show the results in his own presentations. The remaining seven columns in Panel A of Table 1 show that the pattern produced by this procedure is not particularly coherent. We have given each of the rotated factors a label which attempts to summarize, as best as possible, the content of the items that have strong positive loadings (over.5) on the factor. For example, the 'humane society' and 'society where ideas count' items load together positively on the first rotated factor, which we have thus labeled 'better society'. Where possible, the label reflects one of the four subgroups of values identified by Inglehart from Maslow's theory (economic security, physical security, belonging and esteem, and aesthetic and intellectual needs). But, of the four subclusters, only on the second factor do any two of the relevant items load together (the factor is labeled 'belonging', because the 'say in government' and 'say in jobs' items both load on it), and even then one Materialist item, 'maintain order', loads negatively on the same factor. The two prominent items in the first rotated factor are from different

11 274 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH subgroups. 3 None of the other factors produce combinations of variables which make a great deal of intuitive sense. Indeed, on several factors two Materialist items load strongly but with opposite signs. Since it is not absolutely clear from the descriptions in Inglehart (1979) precisely how the items were scored for his own factor analysis of this particular form of the twelve-item battery, an alternative method of scoring the variables was also tried in which the individual items are all scored as dummy variables, coded 1 if the item was rated as one of the two most important goals, in the initial four-item battery, or one of the three most important goals, in the subsequent eight-item battery, and 0 otherwise. Although this method of variable construction is technically less appropriate to subject to factor analytic methods, it produced very similar (but generally slightly stronger) results to the first method (see Panel B of Table 1). For example, the percentage of variance explained by the first factor is 19 percent (compared to 18 percent) and the eigenvalue for the first factor is 2.3 (compared to 2.1). Individual factor loadings are also frequently higher. Somewhat disturbingly, though, the form of the rotated factor matrix varies considerably between the two methods in many details, the most obvious difference being that one produces seven factors and the other six. The best that can be said of the two rotated matrices is that the first factor is similar in each and that they more or less have four factors in common, which can loosely be labeled 'better society', 'free speech', 'defense', and 'environment', although some of these tend to have different baggage attached in the two analyses. Overall the results are not very stable. Next we turn to the equivalent results for the same twelve items measured as ratings. It is generally acknowledged that for most purposes ratings are a technically superior measurement strategy to rankings (Hicks, 1970). The reason rankings are often preferred for measuring priorities of this kind relates to the supposed lack of sufficient variance in ratings, because everyone will value all items highly. As Van Deth (1983^) points out, however, this is an empirical rather than a theoretical question. The means for the twelve items, measured as ratings and rescaled to run from a low of o to a high of 1, range from.63 (for 'beautify cities and countryside') to.92 (for 'fight crime') with an average of.78. The standard deviations range from.15 (for 'fight crime') to.29 (for 'strong defense'), averaging.22. And, as in Van Deth's analysis of Dutch data, panel A of Table 2 shows that the factor analysis of these ratings produces 3 The items theoretically corresponding to the four subdusters are: (economic security) 'fight rising prices', 'economic growth', and 'stable economy 1 ; (physical security) 'maintain order', 'strong defense', and 'fight crime'; (belonging and esteem) 'more say in government', 'more say in jobs', and 'more humane society'; (aesthetic and intellectual needs) 'protect free speech', 'beautify cities and countryside', and 'society where ideas count 5.

12 POL ARI N m H (/) 0 r TABLE I Factor analysis of Materialism-Postmaterialism items based on rankings (Inglehart's method). Principal components analysis with varimax rotation Panel A: Items scored according to rank Rotated factor matrijf First unrotated factor Better society Belonging Free speech Inflation Cntne Defense Environment Materialist items: Maintain order Fight rising prices Economic growth Strong defense Stable economy Fight crime Postmaterialist items: More say in government Protect free speech More say in jobs Beautify cities and countryside More humane society Society where ideas count ~J Jo..11 <n L ~O * *L.10 '4 X a fm > H m90 Z H < mv> T. K) ~J

13 276 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH J 2<g q r r - 01 m WT) q OO o i in O.jl «- o 00 Ko OO I I 1 f I I I I I C100I w Q m 1-1 O OO O O O - I I I irt «««oo] q N q l' ' \ i 000 NO 1 1 o ^o to o^ w r*" i' r r i' r r O -IN M OOO O O O c ON in a SB a 03 CL, a 1 be e O to 8 B ^ J. P 8 2. P ^ S; ti W en c/5 ft. c -a -s T '0 >, 9J 1> O O 3? U S 3 p c.a o o o y u S.9 li

14 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 277 perfectly satisfactory and easily interpretable results, results in fact that are arguably more coherent than those for the rankings in Table 1. In contrast to the Inglehart method, the first unrotated factor this time indicates that all twelve items, both Materialist and Postmaterialist, are positively related to each other. The rotated matrix implies that Materialism and Postmaterialism are two separate dimensions (which in fact correlate positively with each other when constructed as additive scales). The results also appear to be stronger than those in Table 1 in the sense that the first factor explains 28 percent of the variance in the items and has an eigenvalue of 3.4. Furthermore, the pattern in Table 2, Panel A is replicated in the analysis of NSSS data, reported in Panel B. Additive scales for Materialism and Postmaterialism constructed from the ratings have much higher reliabilities than the single twelve-item scale constructed from the rankings (Cronbach's alphas are.68 and.74 for Materialism and Postmaterialism respectively, based on ratings, compared with.53 for the single scale based on rankings). 4 The results in Table 2 indicate that Materialist and Postmaterialist orientations are by no means necessarily mutually exclusive or opposite poles. It is not only theoretically plausible that the same individual might, for example, value both economic security and freedom of expression, the evidence in Table 2 strongly suggests that Materialist and Postmaterialist values do coincide for many individuals. This conceptualization of the theory has important implications for the relative causes and consequences of Materialist and Postmaterialist outlooks. Overall these results are rather unsettling for proponents of a coherent single Materialism-Postmaterialism value dimension. While there is some empirical support for such a dimension it is sustained at only a fairly superficial level. Even the evidence that does support it is relatively weak, with several of the items having uncomfortably small loadings on the first unrotated factor in both panels of Table 1 (as indeed occurs in Inglehart's own analyses for individual countries). Added to this is the fact that the results for the rating items, which suggest that Materialism and Postmaterialism are two largely independent dimensions, are both more robust and more intuitively meaningful overall, as well as being based on a technically preferable measurement procedure (Hicks, 1970; Van Deth, 1983^). 4 In the 1988 welfare survey a third factor having an eigenvalue of just over i.o (the conventional cut-off point) emerged. It appears that because the four items which mate up the standard, abbreviated, Inglehart battery were included separately from the other eight, they tend to load positively together, thus creating a third dimension in the twelve-item factor analysis. However, a number of considerations (including the results from a separate analysis of the eight-item battery and from the NSSS analysis, the eigenvalues of 3.4 and 1.8 for the two factors compared with 1.0 for the third, and the Cronbach's alphas for the two six-item scales) help confirm that the most appropriate solution is the one that produces the two coherent dimensions.

15 278 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH TABLE 2 Factor analysis of Materialism-Postmaterialism items based on ratings. Principal components analysis with varimax rotation. Panel A: National Survey of Australians' Attitudes to Welfare' First Rotated factors* Materialist items: Maintain order Fight rising prices Economic growth Strong defense Stable economy Fight crime Postmatcrialist items: More say in government Protect free speech More say in jobs Beautify cities and countryside More humane society Society where ideas count Panel B: National Survey Social Materialist items: Maintain order Fight rising prices Economic growth Strong defense Stable economy Fight crime Postmaterialist items: More say in government Protect free speech More say in jobs Beautify cities and countryside More humane society Society where ideas count un/uiuicu factor Science Survey 1 First factor i Postmaterialism J L 12. ^ 11 1A Postmaterialism *1 :5 1 ( 1A Rotated factor? Materialism Materialism *1 JO ^ Ratings measured on a five-point scale running from 'extremely important' to 'not at all important'. See text for further details. k Factor loadings of over.5 are underlined. c Ratings measured on a seven-point scale running from 'the most important single thing' to 'a very bad idea'. See text for further details. Sources: National Survey of Australians' Attitudes to Welfare, 1988 (A'=1,814); National Social Science Survey, (A'=3,oi2).

16 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 279 SUBSTANTIVE CONSEQUENCES OF TWO DIMENSIONS RATHER THAN ONE The important question now becomes what are the analytic implications of a two-dimensional conception, as opposed to the traditional one-dimensional view, for the prior causes and behavioral consequences of values? Flanagan has shown, albeit using measures not directly comparable with Inglehart's, that his two independent scales produce 'quite distinct patterns of association' (1982, p. 418). Table 3 sheds further light on this question by showing social structural determinants of Materialism-Postmaterialism scales constructed according to the alternative measurement strategies discussed above. The correlations between the different scales certainly suggest that there is plenty of room for varying results to emerge from this exercise. The single Postmaterialism-Materialism scale constructed from rankings correlates.35 with the Postmaterialism scale constructed from ratings and.46 with the Materialism rating scale. The correlation between the ratings-based Materialism and Postmaterialism scales is -33- The analyses are performed using ordinary least squares regression, with pairwise deletion of missing data. The values scales are recoded to run between o and 1; for those social structural measures that are not dummy variables, scoring is indicated in the table. The first column has the results for the traditional Inglehart scale, which pits Materialists against Postmaterialists based on the ranked items. As the theory would predict, the results emphasize the importance of age, education and religious secularity, among other variables, in shaping Postmaterialist values. Younger people, the better educated, the non-religious and those from fundamentalist Christian or non-christian religions appear more likely to be Postmaterialists than Materialists (as well as part-time workers, the unemployed and people who rent rather than own their homes). The last two columns in Table 3 show a somewhat different pattern. If Materialism and Postmaterialism are simply opposite ends of the same spectrum, we would expect analyses of the two sets of values conducted separately to produce mirror-image results. Using the Inglehart method of course they do (not shown in the table). But using the rating scales, based on the two-dimensional solution in Table 2, we find that on this conception of Materialism and Postmaterialism the two do not have mirror-image determinants. For example, older people are more likely to be Materialist, but age does not affect Postmaterialism at all. Both Materialists and Postmaterialists are likely to have less education than other people (more strongly so for the former). Also, men tend to be less likely than women to be either Materialist or Postmaterialist, although this effect is only statistically significant for the latter. Those without a religion definitely tend not to be Materialists, but such people are only mildly inclined

17 280 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH TABLE 3 Determinants of Materialism-Postmaterialism scales measured by alternative methods: standardized/coefficients regression Male Age (years) Education (years) Income (dollars) Non-manual occupation Work status (ref: full-time) Part-time worker Unemployed Not in labor force Government worker Self-employed Supervisor Rents home Religion (ref: Protestant) Catholic Other religion No religion R 1 Postmat.-Mat. combined ranking scale ** 15** *.11** * * p<.05; ** p<.oi. Source: National Survey of Australians' Attitudes to Welfare, ict88 (#=1,814) **.16**.10 Materialism rating scale beta ** -.20** ** - 15** 15 Postmaterialism rating scale -.09** * ** 04.10**.09* ** 03.06* 05.07* to be Postmaterialists. Likewise, members of minor religions are prone not to be Materialists but have no significant tendency to hold a Postmaterialist orientation. And there are several other similar examples. Based on these results, then, Materialism and Postmaterialism are not mutually exclusive orientations. In fact there is no social structural determinant in this analysis which has an opposite and more or less equal impact on Materialist and Postmaterialist orientations. More broadly, compared to Materialism, Postmaterialism is not well explained at all by social structural variables. Analyses of the NSSS produce broadly similar results (see Bean, 1991, p. 275). The finding that age does not affect Postmaterialism may be considered particularly significant in that it confronts one of the major planks of Inglehart's argument about the 'silent revolution', that successive generations are becoming increasingly Postmaterialist in the late twentieth century. The consequences of education for values are similarly unexpected. What about the behavioral ramifications of values? Perhaps the major reason.06

18 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 281 TABLE 4 Materialism and Postmaterialism scales as predictors of political behavior: regression coefficients* Behavior Support for Australian Community and citizen Democrats contacting activity b beta b beta Materialism rating scale ~-i3.08* Postmaterialism rating scale ** Postmat.-Mat. combined ranking scale.13.09**.20.08** p<.o 5 ; p<.oi. ' Analyses control for all social structural variables listed in Table 3. Source: National Survey of Australians' Attitudes to Welfare, 1988 (N= 1,814). for political scientists being interested in values is that they are thought to influence a range of political behaviors, from political party preferences to participatory activities. Again, the traditional conception of values would suggest that Postmaterialism and Materialism should have opposite influences on political behavior. Yet, elsewhere it has been shown, using the NSSS data, that 'low-level' forms of political protest (such as writing to a newspaper, collecting signatures for a petition or attending a public meeting), are positively related both to Postmaterialism and Materialism, although the effect is only statistically significant for the former (Bean, 1991, p. 277). In Table 4 we show the effects of values on two behavioral variables, support for the Australian Democrats and 'community and citizen contacting activity'. 5 From a theoretical perspective we would anticipate that Postmaterialist orientations would be likely to encourage support for the Australian Democrats, a political party which in recent years has strongly promoted itself as a new politics or environmentalist party (see Papadakis, 1990^). Materialists, on the ' Because few people indicate a preference for the Australian Democrats on any one question (its vote is usually less than 10 percent), Democrat support is measured as a scale combining four separate items (vote in the 1987 House of Representatives and Senate elections, party identification and current voting intention at the time of the survey). 'Community and citizen contacting activity' is also a multiple-item scale comprising questions on whether the respondent has 'worked or cooperated with others in this community to try to solve some of the community's problems', whether the respondent or anyone in the respondent's family has 'ever contacted a government official about some need or problem concerning the community as a whole' and whether the respondent or anyone in the respondent's family has 'ever contacted a government official to seek help with a personal problem'. These three items form a clear scale, even though they cover two slightly different dimensions of citizen participation as identified by Verba et al. (1978), 'communal activity' and 'particularized contacts'. Both dependent variables are rescaled to run between 0 and 1.

19 282 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH other hand, would if anything tend not to support the Democrats, but the effect may not be at all large. With respect to community-oriented political activity, we would again expect to find a positive effect from a Postmaterialist outlook (given that Postmaterialist values emphasize citizen participation) while Materialism may not be related to such activities either one way or the other. The first two rows in Table 4 contain the regression effects, on each of the two dependent variables in turn, of the Materialism and Postmaterialism scales constructed separately from the ratings data. In each case, both scales are of course entered in the same equation, so the effects are net of each other (and the analyses also control for all the social background variables in Table 3). Focusing initially on the Australian Democrats, we find that Materialist values are negatively related to support for the party, but Postmaterialism has no significant effect on Democrat support. As well as providing a further indication that Materialism and Postmaterialism are not simply direct opposites, the fact that this result is counter-intuitive makes it all the more noteworthy. For community and citizen contacting activity, the results are more in line with expectations: Postmaterialism has the anticipated positive effect, while Materialism has a negative but non-significant coefficient. The third row in table 4 rounds out the analysis by showing how a single Postmaterialism-Materialism scale, based on rankings, would lead to the interpretation that both of the two dependent behaviors are distinguished by a direct contrast between Materialism and Postmaterialism, in which Postmaterialists tend to engage in these activities as opposed to Materialists who do not. VALUE CHANGE AND DOMINANT VERSUS ALTERNATIVE PARADIGMS Claims of a single polarized value dimension which sees Materialists at one end lined up against Postmaterialists at the other extreme are at variance with much of the empirical evidence and indeed even with some of the basic tenets of the theory of value change as elaborated by Inglehart himself. The challenge of explaining this apparent inconsistency has been taken up by Cotgrove and Duff (1981), who have argued that different values are relevant to different contexts. This implies a modification or refinement of the notions of Materialism and Postmaterialism. Cotgrove and Duff have distinguished, for example, between values held in public and in private: 'the goals which individuals seek to maximise for society may not correspond precisely to the hierarchy of values which operates in their personal lives'

20 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 283 (1981, p. 101). They have argued that values will be influenced by different stages during the life cycle and that their relevance will depend on particular contexts. It may therefore be misleading to focus on apparent inconsistencies in the behavior of Materialists and Postmaterialists. For example, 'It is quite consistent to act according to material values in an occupation (to be dissatisfied with the salary, for example), and at the same time give low priority to economic growth as a goal for society' (Cotgrove and Duff, 1981, p. 101). Marsh (1975) has come to a similar conclusion about the variation in the values of the same individual in the public and private domains. The idea of a shift from Materialist to Postmaterialist values has influenced debates about the new politics and new social movements and been taken up by several writers. Rather than focus on Materialism and Postmaterialism, these writers have suggested a contrast between a dominant and an alternative social paradigm. The term paradigm is derived by Cotgrove (1982) from the work of Kuhn (1970). For Cotgrove paradigms 'provide maps of what the world is believed to be like'. Paradigms 'constitute guidelines for getting about and for identifying and solving problems' and provide 'the framework of meaning within which "facts" and experiences acquire significance and can be interpreted' (1982, p. 26). Cotgrove has argued that in comparing the beliefs and values of industrialists (in other words, the dominant groups in industrial society) and of environmentalists we are confronted with two contrasting perspectives on what ought to be done about solving societal problems. The dominant paradigm shares many of the characteristics of a Materialist orientation. The alternative paradigm appears to overlap with Postmaterialism. In addition, the alternative paradigm appears to represent many of the values espoused by environmentalists and by new social movements. Like Inglehart, those who have written about an alternative environmental paradigm posit a fundamental conflict with the dominant paradigm. Milbrath (1984, p. 99) has written about the displacement of economic values by a 'wider range of quality of life values'. One objection to these accounts is their apparent failure to examine the inconsistency between values and behavior. Someone may claim to be a Postmaterialist (and vote for a Green Party), but will behave like a Materialist (by fighting for an improved wage and, at least implicitly, supporting more economic growth). However, writers like Cotgrove have recognized that the notions of a dominant and alternative paradigm only provide a rough guide to points of potential conflict. Most people are neither outright Materialists nor Postmaterialists; they cannot easily be grouped as proponents/opponents of the dominant/alternative paradigm. Most people have values and beliefs that correspond to both dominant and alternative paradigms.

21 284 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH CONCLUSIONS The notion of a single polarized dimension of conflict is at odds with much of the evidence about the recent development of environmentalism (see Papadakis, 1993). Throughout the world established organizations have adapted to many of the challenges posed by environmentalism. For instance, arguments for sustainable development in the 1990s represent a major effort by governments and other established organizations to provide an alternative to the focus on development versus environmental protection (see World Commission on Environment and Development, 1990). Writers like Cotgrove (1982, p. 91) have recognized that environmentalists and other social actors have tended to draw on both the dominant and alternative social paradigms. The distinction between the two paradigms therefore serves mainly as a heuristic device and thus has limited application. By contrast Inglehart has insisted that, because there are significant differences in the political behavior of Materialists and Postmaterialists, he has identified a single, basic conflict dimension in society. Of course social actors have frequently posited a fundamental conflict between Materialism and Postmaterialism or between economic growth and environmental protection in order to dramatize the differences in their policy platforms and to mobilize popular support. In a sense Inglehart has done a similar thing. However, only small groups of people can be identified as strong supporters of a dominant or alternative paradigm. In addition, environmentalists are not necessarily deep ecologists, nor are they all militant activists challenging every aspect of the institutionalized order. Similarly 'non-materialists' are concerned about improving the quality of life in a variety of ways: some have emphasized democracy, others have been mainly concerned about environmental protection and others have been preoccupied with both. The emphasis on Materialism and Postmaterialism or on a dominant versus an environmental paradigm is useful insofar as it represents significant aspects of the ideological struggle for social and political change. However, this is a far cry from arguing for a single conflict dimension over Material and Postmaterialist values. We have shown that it is difficult to sustain such a claim empirically and that a multi-dimensional treatment of Materialist and Postmaterialist values has important implications for our understanding of the causes and consequences of these phenomena. Perhaps the most useful purpose served by the notions of a dominant and an alternative paradigm and of Materialism and Postmaterialism is that they at least enable us to locate and to draw attention to some of the significant shifts in public preferences and in policies over the past few decades.

22 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 285 APPENDIX Extract from questionnaire, showing format of Materialist-Postmaterialist battery 52. And now some questions about the country as a whole. There is a lot of talk these days about what the aims of Australia should be for the next 10 years. Here is a list of some of the aims that different people would give top priority. How important do you think it is to Maintain order in the nation 2. Give people more say in important government decisions 3. Fight rising prices 4. Protect freedom of speech Extremely important Very important 1 2 Fairly important Not important Not at all important 5 If you had to choose among these four aims, which would be your first choice? Which would be your second choice? Which would be your third choice?... (Place the number of the item in the appropriate boi.) First choice Q Second choice Q Third choice 53. Below are some of the aims people say our country as a whole should concentrate on. How important do you think it is to Maintain a high rate of economic growth a. Make sure that this country has strong defense forces 3. See that people have more say about bow things are done at their jobs and in their communities 4. Try to make our cities and countryside more beautiful 5. Maintain a stable economy 6. Progress toward a less impersonal and more humane society 7. Progress toward a society where ideas count more than money 8. Fight against crime Extremely tmportant Very tmportant Fairly important Not important Not at all important Of course, all of these are important to us in one way or another, but which three are most important to you personally? (Place the number of the item in the appropriate bo%.) Most important Q Second most important Q Third most important Q

23 286 INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH REFERENCES Baker, Kendall L., Dalton, Russell, and Hildebrandt, Kai (1981): Germany Transformed, Cambridge, Harvard University Press. Bean, Clive (1991): 'Participation and Political Protest: A Causal Model with Australian Evidence', Political Behavior, 13, Bernstein, Eduard (1899): Die Voraussetzungen des Sozialismus und die Aufgaben der Sozialdemokratie, Stuttgart, J. H. W. Dietz (translated by E. C. Harvey, 1961, Evolutionary Socialism: A Criticism and Affirmations, New York, Schocken Books). Boltken, Ferdinand and Jagodzinski, Wolfgang (1985): 'In an Environment of Insecurity: Postmaterialism in the European Community, 1970 to 1980', Comparative Political Studies, 17, Butler, David and Stokes, Donald (1974): Political Change in Britain, 2nd edn., London, Macmillan. Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E. (i960): The American Voter, New York, John Wiley and Sons. Clarke, Harold D. and Dutt, Nitish (1991): 'Measuring Value Change in Western Industrialized Societies: The Impact of Unemployment', American Political Science Review, 85, Cotgrove, Stephen (1982): Catastrophe or Cornucopia, Chichester, John Wiley and Sons. Cotgrove, Stephen and Dutt, Andrew (1981): 'Environmentalism, Values, and Social Change', British Journal of Sociology, 32, Dalton, Russell J. (1981): 'The Persistence of Values and Life Cycle Changes', Politische Vterteljahresschrift, 22(Sonderheft 12), Dalton, Russell J. (1988): Citizen Politics in Western Democracies, Chatham, New Jersey, Chatham House Publishers. Flanagan, Scott C. (1982): 'Changing Values in Advanced Industrial Societies: Inglehart's Silent Revolution from the Perspective of Japanese Findings', Comparative Political Studies, 14, Flanagan, Scott C. (1987): 'Value Change in Industrial Societies', American Political Science Review, 81, Hellevik, Ottar (1993): 'Postmaterialism as a Dimension of Cultural Change', International Journal of Public Opinion Research, 5, Hennessy, Bernard (1985): Public Opinion, Monterey, CA, Brooks/Cole. Herz, Thomas (1979): 'Der Wandel von Wertvorstellungen in westlichen Industriegesellschaften', Ko'lner Zeitschrift fiir Soziologie and Sozialpsychologie, 31, Hicks, Lou E. (1970): 'Some Properties of Ipsative, Normative, and Forced-Choice Normative Measures', Psychological Bulletin, 74, Inglehart, Ronald (1971): 'The Silent Revolution in Europe: Intergenerational Change in Post-Industrial Societies', American Political Science Review, 65, Inglehart, Ronald (1977): The Silent Revolution, Princeton, Princeton University Press.

24 POLARIZED PRIORITIES OR FLEXIBLE ALTERNATIVES? 287 Inglehart, Ronald (1979): 'Value Priorities and Socio-Economic Change'. In Samuel H. Barnes, Max Kaase et ai, Political Action, Beverly Hills, Sage. Inglehart, Ronald (1981): 'Post-Materialism in an Environment of Insecurity', American Political Science Review, 75, Inglehart, Ronald (1990): Culture Shift in Advanced Industrial Society, Princeton, Princeton University Press. Jackson, David J. and Alwin, Duane F. (1980): 'The Factor Analysis of Ipsative Measures', Sociological Methods and Research, 9, Kelley, Jonathan, Cushing, Robert G., and Headey, Bruce (1987): Australian National Social Science Survey, 1084: Users Guide, Canberra, Social Science Data Archives, The Australian National University. Kuhn, Thomas S. (1970): The Structure of Scientific Revolutions, 2nd edn., Chicago, Chicago University Press. Lafferty, William M. (1975): 'Basic Needs and Political Values: Some Perspectives from Norway on Europe's "Silent Revolution"', Ada Sociologica, 19, Lafferty, William M. and Knutsen, Oddbjorn (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, Marsh, Alan (1975): 'The Silent Revolution, Value Priorities, and the Quality of Life in Britain', American Political Science Review, 69, Marsh, Alan (1977): Protest and Political Consciousness, Beverly Hills and London, Sage. Maslow, Abraham K. (1954): Motivation and Personality, New York, Harper and Row. Milbrath, Lester W. (1984): Environmentalists: Vanguard for a New Society, Albany, State University of New York Press. Muller-Rommel, Ferdinand (1983): 'Die Postmaterialismusdiskussion in der Empirischen Sozialforschung: Politisch und wissenschaftlich uberlebt oder noch immer zukunftsweisend?', Politische Vterteljahresschrift, 24, Papadakis, Elim (1990a): Attitudes to State and Private Welfare: Analysis of Results from a National Survey, Social Policy Research Centre, Reports and Proceedings No. 88, Sydney, Social Policy Research Centre, University of New South Wales. Papadakis, Elim (1990^): 'Minor Parties, the Environment and the New Policies'. In Clive Bean, Ian McAllister and John Warhurst (eds.), The Greening of Australian Politics, Melbourne, Longman Cheshire. Papadakis, Elim (1991): Does the New Politics Have a Future?'. In Francis G. Castles (ed.), Australia Compared, Sydney, Allen and Unwin. Papadakis, Elim (1993): Politics and the Environment: The Australian Experience, Sydney, Allen and Unwin. Reimer, Bo (1988): 'No Values New Values? Youth and Postmaterialism', Scandinavian Political Studies, 11, Van Deth, Jan W. (1983a): 'The Persistence of Materialist and Postmaterialist Value Orientations', European Journal of Political Research, 11, Van Deth, Jan W. (1983*): 'Ranking the Ratings: The Case of Materialist and Post- Materialist Value Orientations', Political Methodology, 9: