1 Introduction

In this article we study the stratification of cultural consumption in Israel with two emphases. First, the distinction between class and status as representing two different sources of influence: the economic aspect of social position cast in terms of a Weberian conceptualization of class and the esteem aspect of social position cast in terms of a Weberian definition of status. Second, we emphasize dimensions of stratification that are salient in the Israeli context, namely ethnicity and religiosity.

1.1 Class and status

Traditionally, research on the stratification of cultural consumption and taste tended to emphasize the association between class and cultural consumption and preferences (Bourdieu 1984). However, within the Weberian theoretical framework, status situation and the esteem attached to it is a dimension that is associated with, but analytically separate from, class (Weber 1968 [1922]).Footnote 1

Accordingly, it is expected that status will be more important than class in influencing cultural consumption because consumption is an aspect of lifestyle used to demarcate status group boundaries. Status groups reproduce themselves and legitimize their privileged position by adopting distinct cultural preferences (Lamont and Molnár 2002). The consolidation of cultural preferences into distinguishable patterns of taste has been a foremost concern in the research of Veblen (1899); Levine (1988); Bourdieu (1984), and others, who have shown how cultural consumption serves as a means for creating group boundaries. Still, most research in this field has regarded class as the main facilitator of cultural boundaries. Hardly any research has offered an operationalization of status, let alone contrasted it with class in examining cultural consumption patterns. In this article, we examine the relative contribution of class and status in shaping cultural consumption patterns in Israel.

Previous research has already shown the significant role that various factors have in influencing cultural consumption. These include, for example, income and employment characteristics, education, gender, race, urban status, age, family structure, and parental cultural capital (Lamont and Molnár 2002; Katz-Gerro 2004). The focus of this article is on the effect of status and class, net of these factors, with the expectation that status will be a more important determinant than class. But, when studying Israeli society, one must consider two additional elements that play a central role in shaping group boundaries, namely ethnicity and religiosity.

1.2 Ethnicity and religiosity in the Israeli context

Research on Israeli society is usually cast in terms of ethnicity, religiosity, and socio-economic characteristics. As a consequence, very few studies consider the role of social class as a dimension of differentiation and stratification. The economic dimension of stratification is usually conceptualized as “socio-economic”, meaning some combination of occupation, prestige, and, where available, income. Many researchers subscribe to the view that although economic inequality is wide and increasing, classes are very weakly structured in Israel. Class-based political ideologies and political recruitment are not prevalent, the link between class position and political inclination is weak (Horowitz and Lissak 1990), and interclass mobility is by no means closed off (Yaish 2004). Moreover, Horowitz and Lissak (1990) maintain that the privileged classes do not exhibit distinctive cultures and lifestyles.

This notwithstanding, the general picture that emerges from the very few studies on the social bases of cultural consumption that have been carried out in Israel is that class and education are among the main determinants of patterns of consumption and taste. No research in Israel has explored the possibility that status affects cultural stratification. The only exception is our own analysis (Katz-Gerro et al. 2007) reporting that in Israel musical tastes are shaped by father’s status. The few studies on cultural stratification are concerned with the stratification of cultural consumption and tastes (Katz and Gurevitz 1973; Katz 1992; Katz-Gerro and Shavit 1998); musical tastes and ethnicity (Benski 1989); leisure and social networks (Ginsberg 1975); lifestyle and ethnicity (Arian and Talmud 1991); the characteristics of lifestyles (Goldman and Kipnis 2003); and sport and gender (Shamir and Ruskin 1983).Footnote 2

To understand cultural stratification in Israel one must recognize several deep cleavages running across Jewish Israeli society (Horowitz and Lissak 1990).Footnote 3 First, Israel is an ethnically stratified society. The main ethnic division in the Jewish population divides Mizrachi (Mizrachim in the plural) Jews (of Asian or North African origin) and Ashkenazi (Ashkenazim in the plural) Jews (of European or American origin) into two roughly equally sized groups. In addition to these two main Jewish ethnic groups, there exists a group of third-generation Israelis (i.e., Israeli born to Israeli born parents). Third-generation Israelis are more similar to the Ashkenazi than to the Mizrachi, in terms of educational attainment, occupational attainment, and cultural preferences (Cohen et al. 2007; Katz-Gerro et al. 2007). The ethnic hierarchy in Israel is in transition. While the Ashkenazi were the dominant group since the inception of the state and played a central role in many domains, third-generation Israelis are becoming a more important group. The ethnic cleavage in Israel is unique because it originates in the historical spread of Jews around the world, and because it is based not on national identity but rather on differences in religious rites and customs as well as on territorial differences.

Following the end of World War II and into the 1950s, Jews the world over migrated to Israel. Immigrants from different countries had varying cultural, educational, professional, and demographic characteristics. Accordingly, their ability to use these resources affected their life trajectories (Horowitz and Lissak 1990). In contemporary Israel, Ashkenazim are over-represented in positions of political power and in the professions and have higher incomes. On average, Mizrachim are less educated than Ashkenazim and are typically concentrated in lower white-collar and skilled blue-collar occupations (Smooha 1993; Yaish 2004). Ethnicity is central in determining the contours of Israeli society and is present in all of its institutions (e.g., the educational system, the political system, the military, the labor market, and the cultural field). Significant differences exist between ethnic groups in various aspects of the lifestyle, including food, esthetic preferences, artistic canons, and the like. In addition, Ashkenazi Jews are over-represented within producers and consumers of legitimate culture (Regev and Seroussi 2004).

When Israel became an independent state in 1948, and for a long time thereafter, it was ruled by a cultural hegemony of the “veteran residents” who had arrived in Israel at the time of the British Mandatory regime in Palestine. These veterans were from East European countries, and their hegemony—with its Western, modern, and secular cultural preferences—had to face up to the challenge posed by the large waves of immigration that entered Israel from the Muslim states in the 1950s, and which threatened to generate radical changes in the social and cultural order. Affording recognition to these various ethnic groups and their cultural heritage was inconsistent with the ideology of the “melting pot”, a national project that fostered the creation of a homogeneous Jewish public, based on the negation of the Diaspora and the creation of the “new Israeli”. According to this ideology, the Mizrachim were perceived as a group that had to be assimilated (Shenhav 2003). In this context, Regev and Seroussi (2004; Regev 2000) use the term “national cultural capital” to refer to cultural capital and habitus that define membership in a specific national culture. In the Israeli case, national cultural capital is considered to be part of a wider legitimate culture, which is Ashkenazi culture. Therefore membership in the collective is conditioned by competence in Ashkenazi culture.

Over the years, and especially since the early 1980s following major political transformations, the old hegemony began to lose its power. Mizrachi leaders increasingly voiced opinions that challenged homogeneous Israeli cultural identity, and a significant wave of immigration from the former Soviet Union in the early 1990s transformed the very structure of ethnicity in Israeli society (Kimmerling 2004). Nevertheless, significant differences still exist between Ashkenazi, Mizrachi, and third-generation Israelis.

The second main cleavage divides Jews in Israel according to religious observance: between the ultra-orthodox, the “modern” orthodox, the traditional, and the secular (Smooha 1993). The cleavage between secular and religious Jews in Israel is not dichotomous. It is best represented as shades on a continuum that can be cut in different places according to different criteria (Horowitz and Lissak 1990). These criteria mainly revolve around voting for religious political parties, sending children to religious schools, and the degree of adherence to Jewish laws. Degrees of religiosity are associated with degrees of adherence and observance of the laws, a cultural lifestyle, a tendency for social and ecological isolation, and degree of commitment to the national cultural center. For example, all religious Jews refrain from traveling on Shabbat (Saturday, the Jewish sabbatical day). The ultra-orthodox do not listen to female singing voices, and abstain from watching television. In some religious groups, men and women do not mix in public places. Meanwhile, some traditional Jews may travel on Shabbat, while strictly adhering to Jewish dietary laws. Differences in position on the religious-secular continuum are also reflected in the degree to which social networks are open. For example, most ultra-orthodox men study religion rather than work, a practice that isolates them from much of society.

Of the roughly 20% of the Jewish population who are considered religious, the majority are well-integrated into Israeli society in terms of residential patterns, participation in military service, and patterns of labor force participation. Approximately one-third of the religious are ultra-orthodox. Members of ultra-orthodox communities live in largely segregated neighborhoods, and their children attend schools belonging to a separate educational system whose curricula contain few or no secular subjects. Ultra-orthodox men are generally exempted from mandatory 3-year military service after registering their religious studies at a yeshiva (a post-secondary religious educational institution) as their full-time vocation. These men have very low official rates of labor force participation as they continue studying at the yeshiva for a substantial part of their lives.

Political, economic, and social conflicts are prevalent among the secular, religious, and ultra-orthodox segments of the population. For example, religiosity in Israel is considered an important determinant of political attitudes (Shamir and Arian 1999), educational and occupational attainment, and several other life outcomes. Family size, as well as the likelihood of being below the poverty line, is highly associated with religiosity, particularly among the ultra-orthodox (Stier and Lewin 2002). The different groups also practice quite different cultural lifestyles. There are differences in the amount of time spent on different activities such as eating and leisure activities. The ultra-orthodox refrain from watching television, or going to the movies because such activities diverge from a strictly observant lifestyle (Katz and Gurevitz 1973).

A discussion of cultural consumption in Israel should also acknowledge the relative insignificance of place of residence in influencing consumption opportunities. First, Israel is a small country. With a population of seven million, and an area slightly smaller than New Jersey, distance from urban areas has only a moderate impact on access to culture (Katz 1992; Katz-Gerro 2002). Second, although most cultural production occurs in the city of Tel Aviv, there is a well-developed system of cultural dissemination into the periphery, in particular music, theater, other shows, and cinema. The distribution of cultural activities to the periphery is partly supported by state agencies. To sum, membership of an ethnic or a specific religious group in Israel implies belonging to a social community with particular cultural norms. There are important correlations between religiosity and ethnicity in Israel. Mizrachi Jews are more likely to identify as traditional or religious (Ayalon et al. 1991). Nevertheless, previous research on religiosity, ethnicity, and leisure time use has demonstrated that these dimensions are distinct (Sullivan 1998). The effects of this membership operate through factors such as family, and the power of peer and community influence. In particular, religious beliefs are part of a social system that shapes and places strict limits on cultural preferences and cultural behavior. In this article we explore the way in which these unique characteristics of Israeli society may affect cultural consumption patterns.

1.3 Cultural consumption patterns

In the past 15 years research on cultural stratification has emphasized a phenomenon that was termed cultural omnivorousness. Studies conducted in various national contexts have reported that members of the upper strata increasingly exhibit a preference for both highbrow and lowbrow cultural repertoires. In the most influential work in this field, Peterson and colleagues (Peterson and Simkus 1992; Peterson and Kern 1996) studied musical tastes in the USA, comparing data from 1982 and 1992. They showed that compared with findings in the 1982 data, in 1992 individuals of more advanced social background did not restrict their taste to highbrow music genres, and were consequently described as cultural omnivores. Peterson has articulated the omnivore thesis, which argues that individuals of high social standing will distinguish themselves by adopting a wide variety of cultural competencies. Extensive research has reported similar findings in other countries (Peterson 2005; Sullivan and Katz-Gerro 2007), but no research has been conducted on cultural omnivorousness in Israel.

In this article we test hypotheses concerning the relationship between consumption and the social structure, while adhering to the distinction between the effects of class and status. We expect that status rather than class will be the main determinant of cultural consumption patterns. In addition, as we explained in the previous section, we introduce and elaborate on two dimensions that are important for understanding cultural consumption patterns in the Israeli context, namely, ethnicity and religiosity. We expect that both will be important for shaping cultural consumption, but the lack of research on these dimensions prevents us from formulating any directional hypotheses.

2 Data and variables

2.1 Data

We employed two data sets for this study. The first is the 1995 census data set on detailed occupations of married couples, which was used to construct the status order for Israeli society. The measure that we adopt here follows work by Chan and Goldthorpe (2004), builds on Laumann’s (1966) differential association approach, and defines status in relation to social distance. In constructing the Israeli status scale we rely on the notion that patterns of intimate relationship indicate proximity among occupations, and can be used to generate a ranking of occupations that represents their relative status standing. We consider marriage as intimate relationship and examine the similarity of husbands’ and wives’ occupations in order to translate similarity and difference to social distances among occupations.

The data set included 211,724 households of whom 177,593 households entered our analyses. Based on a two digit classification of the Israeli Bureau of Statistics, we constructed 33 occupational categories (S. Raz et al. 2008, Unpublished manuscript). We consider current, or if unemployed, last job in the previous year. Unemployed for whom we had no information on current/last occupation were assigned to a separate category. Similarly, those who were not part of the Israeli labor force were assigned to a separate category. Indices of dissimilarity between each pair of the 33 occupational categories of married couples were used as an input in a multidimensional scaling analysis to generate a map of proximity between occupations. From this analysis a two-dimensional map emerged. One dimension can be interpreted as representing gender occupational segregation. The other dimension represents distribution of occupations along several axes: manual versus non manual, white collar versus blue collar, and working with things versus working with people. This distribution of occupations points to a hierarchy of occupations. Since this hierarchy is based on intimate associations, it indicates social distance, and we interpret it as representing a status order. We conducted analyses to determine whether the same hierarchical dimension applies to different sub-groups in the population: Jews and Arabs, Mizrahi and Ashkenazi, immigrants after 1989 and veterans. Once this dimension was validated we used it to construct a status order of occupations.Footnote 4 A detailed list of major occupations, status scores, and relative group sizes appears in Appendix A in a previous paper (Katz-Gerro et al. 2007).

The second data set is based on two surveys on cultural consumption conducted in 2006 (N = 808) and 2007 (N = 401). The surveys were conducted via telephone interviews with a nationally representative sample of the Israeli Jewish population.Footnote 5 The surveys were specifically designed by the authors for this project, and were carried out by the survey research facilities of the University of Haifa by means of a computerized telephone interview system.

2.2 Variables

The dependent variables pertain to six questions on the consumption of cultural activities outside the home. On a scale of one to five (1—never, 2—once or twice, 3—three or four times, 4—once in 2 months, 5—at least once each month), respondents were asked to report frequency of participation during the last 12 months for the following activities: going to the cinema, going to the theater, visiting a museum or a gallery, attending a classical music concert or opera, attending another kind of musical show, going to the ballet or a dance performance. The distribution of all six items is skewed, with a large percentage of respondents who had not participated in any of those activities during the 12 months prior to the survey. Therefore, we use the variables in a dichotomous form, indicating participation or non-participation. Table 1 lists the activities in order of popularity. Going to museums or galleries is the most popular activity (60% attendance), followed closely by going to the movies (58% attendance), visiting the theater (54% attendance), and attending musical shows that are not classical music or opera (43%). The two least frequently attended activities, far behind the other four, are going to a classical music concert or an opera (20%) and going to a ballet or dance performance (17%). Note that all the analyses presented here pertain to N = 850, which is the number of respondents for whom we have information available on all variables.

Table 1 Percent attendance in cultural consumption activities in the last 12 months
Full size table

Our first objective was to identify consumption patterns. To do this we applied latent class analysis. The results are reported in Tables 2 and 3. Table 2 shows fit statistics results for two-, three-, and four-class solutions. None of these solutions reaches the standard (p < 0.05) statistical significance level. Nevertheless, the four-class model significantly improves the fit over the three-class model (∆G2 = 26.5 with ∆df = 7; p < 0.05). Although we are unable to use the results of this analysis, we can gain some insights from an inspection of the pattern that emerges.

Table 2 Latent class analysis of indicators of cultural consumption
Full size table
Table 3 Relative size of four latent classes, and conditional probability of attending each cultural event for each class
Full size table

The results of a four-class solution, shown in Table 3, suggest that consumption patterns in Israeli society form four clusters along the active–inactive dimension that should be interpreted as discrete categories. The first class represents respondents inactive on the indicators studied (the “inactives”), while the remaining three classes represent those who are active (the “actives”). Within the active classes, classes 3 and 4 represent the active omnivores who participate in all six activities. The main difference between individuals in these two classes lies in their level of activity. Whereas those in class 4 participate in all six activities with relatively high probability, those in class 3 participate in these activities with much less probability. Finally, class 2 represents individuals who mostly participate in popular activities. We use those insights suggested by latent class analysis to construct four consumption profiles: inactives, non-highbrow actives, semi-omnivores, and omnivores.

Inactives are respondents who had not engaged in any of the six activities in the preceding year. Non-highbrow actives are respondents who attended one to four activities during the year, excluding ballet/dance or classical music concert/opera. Semi-omnivores are respondents who attended one to four activities in the preceding year, where at least one of those was ballet/dance or classical music concert/opera. Finally, omnivores are respondents who attended any five or six activities (meaning at least two highbrow activities). The last two groups, semi-omnivores and omnivores, are quite similar, and one might ask why they should be kept separate. Results of the preferred latent class solution produced four classes and suggested that two of them are omnivorous classes that can be distinguished by their degree of omnivorousness. In constructing our consumer types, we relied on this result by separating omnivorousness into two levels.

Three important qualifications to these labels are required. First, individuals labeled inactive may well be very active in other cultural consumption activities not recorded here (e.g., live performance shows, cultural festivals) or in other domains, such as sports, outdoor leisure activities, and the like. Second, the differentiation between highbrow and lowbrow among the six items is not straightforward. On the one hand, cinema is an activity that in many countries is considered lowbrow, but in the Israeli context it could refer to both highbrow and lowbrow tastes. On the other hand, we can point to four items included in our data that are regarded in the literature as representing highbrow cultural consumption: ballet/dance, classical concert/opera, theater, and museum/gallery. However, in the Israeli context, theater and museums/galleries can indicate both highbrow and lowbrow activities. This is because individuals, who attend children’s museums, popular science museums, or social activities taking place inside museums, will be wrongly classified as highbrow, compared to individuals who attend art museums or galleries. Similarly, individuals who attend comedy and light entertainment shows in theater halls will be wrongly classified as highbrow compared to those who attend theater shows. Therefore, we can point to only two items—ballet/dance and classical concert/opera—that can truly be considered cultural activities associated with highbrow taste. For these reasons we labeled the second type non-highbrow actives rather than lowbrow actives, and as a result we are unable to fully test the homology thesis. Third, the use of the term omnivore here somewhat diverges from the original definition proposed by Peterson and others (Peterson 2005). Peterson and Kern (1996) defined omnivores as individuals who, in addition to liking highbrow musical genres, also like middlebrow and lowbrow musical genres. Since we are unable to determinately identify clear lowbrow activities in our data, for reasons explained above, the term omnivore in our research describes respondents who have eclectic cultural consumption preferences that combine highbrow and non-highbrow activities.

Table 4 presents the distribution of the consumption types in two forms. First, we characterize the division between inactive respondents (who did not attend any of the activities) and active respondents (who attended at least one activity). Second, the active respondents were further differentiated according to the types described above. Inactive respondents constitute 15.6% of the sample, and the actives are 84.4%. Among the actives, omnivores and semi-omnivores constitute about 15% each, while the non-highbrow actives account for about 55%.

Table 4 Distribution of cultural consumption types
Full size table

Table 5 presents descriptive statistics of the independent variables. Education is measured in years, as well as age. Religiosity is measured on a five-point interval scale, including 1—secular, 2—secular who keeps some of the Jewish laws, 3—traditional, 4—religious, 5—orthodox. Ethnicity is a variable that combines respondents’ and their fathers’ country of birth. It is represented by three dummy variables, for second-generation Israelis, Israelis of American-European origin (Ashkenazi), and Israelis of Asian-African origin (Mizrachi). Income pertains to net household income, coded to ten categories.Footnote 6 We use the natural log of the midpoint of each category. Class categories are based on a four-category version of the CASMIN class schema (salariat class I + II, intermediate class IIIa + V, petty bourgeoisie IVabc, and working class VIIab + IIIb + VI). Finally, the status-order variable was as described in the previous section.

Table 5 Means (Standard deviations) and proportions of independent variables
Full size table

3 Results

Table 6 presents descriptive information on the consumption types in reference to the main stratification dimensions. On average, actives are more educated than non-actives, their mean status score is much higher, and so is their income. Comparing non-highbrow actives, semi-omnivores, and omnivores, we can see a gradual rise in years of education, status score, income, percent of Ashkenazim, and percent of upper class respondents, and an increase in average level of religiosity. This means that individuals in privileged social groups tend to be both more active in cultural participation and more omnivorous.

Table 6 Means (Standard deviations) and proportions of selected socio-demographic and stratification variables of cultural consumption types
Full size table

So far, we have evidence of a clear pattern of association between stratification variables, in particular education, income, class, and status, and cultural consumption profiles. The next step is to apply a multivariate analysis. To this end we apply multinomial logistic regressions to the consumption-types variable. In Table 7 we present likelihood ratio tests for two nested models. Model 1 includes all of the socio-economic background variables and status score. Model 2 includes all of the variables from model 1, and adds household income and respondent’s class. In model 1 all the variables have significant effects, including status score. In model 2 all the variables remain significant except for respondent’s status score.Footnote 7

Table 7 Likelihood ratio tests for terms in the multinomial logit models on four cultural consumption types (N = 850, Nagelkerke R 2 = .24)
Full size table

The finding here is that while status is significantly associated with cultural consumption, when controlling for class and income the effect of status becomes insignificant.Footnote 8 The immediate implication is that material bases of stratification, class, and income are more important than the esteem aspect measured by status. This result also clearly indicates that cultural consumption is socially stratified by class, but also by ethnicity, religiosity, and education. To shed light on the applicability of the omnivore thesis to Israeli cultural stratification we need to evaluate the detailed associations between social stratification variables and cultural consumption types, which we present in Table 8.

Table 8 Estimates (and standard errors) of logistic regression models predicting cultural consumption
Full size table

Table 8 displays the parameters associated with model 2 of Table 7. It shows six different contrasts among the four cultural consumption types in models that predict the probability of being classified into each type, conditional on age, gender, ethnicity, religiosity, education, status, class, and household income. The reference group in the first three contrasts is the omnivore type. The general pattern that emerges here is that socio-demographic and stratification measures affect the likelihood of belonging to the inactive and non-highbrow actives as against the omnivores, and only marginally affect the likelihood of being a semi-omnivore rather than an omnivore.

Inactives tend to be Mizrachi, more religious, less educated, more often male, have a lower income and belong to the petty bourgeois (as opposed to salariat) than omnivores. Non-highbrow actives tend to be Mizrachi, younger, more often male, and from the intermediate class (compared with salariat) than omnivores. By contrast, the omnivores are more educated, less religious, more often Ashkenazi or second-generation Israeli, more affluent, and from the salariat class. Omnivores and semi-omnivores are mainly differentiated by class, such that in the latter group fewer respondents are from the salariat class. Interestingly, omnivores are not differentiated by status position, which indicates, as we have already seen in Table 7, that in Israel status is less influential than class and income in determining cultural repertoires. In the discussion section we will further elaborate on these conclusions.

The remaining three columns in Table 8 show the other possible contrasts among the four cultural consumption types. When the inactives are contrasted with any of the three active groups (columns 1, 4, and 6) it is evident that they tend to be Mizrachi, more religious, less educated, and with lower household income.Footnote 9 Non-highbrow actives are likely to be Mizrachi, younger, less educated, and more often males than semi-omnivores.

To sum, we emphasize three conclusions. First, all types of cultural consumption analyzed here are conditioned by various economic and demographic dimensions of stratification that are central in the Israeli context. Second, material bases, namely class and income, are more salient than the esteem aspect of social position, represented by status, in shaping types of cultural repertoires. Finally, the typology of cultural types presented here shows the importance of socio-demographic attributes, economic characteristics, and other measures of stratification in differentiating cultural activities along two main axes. The first is a distinction between the actives and the inactives. The second is a distinction between the two types of omnivores and the non-highbrow actives, that is, between those who, among other things, engage in highbrow activities and those who do not. The final conclusion could not have been reached had we used a simple one-dimensional measure of cultural activities.

4 Discussion

In this section we discuss two main conclusions that emerge from the analyses. First, class is more important than status in shaping cultural consumption in Israel. In what follows we discuss possible reasons for this finding. Second, individuals from more advantageous social positions are more likely to adopt omnivorous cultural consumption practices. This is not a surprising result and it joins numerous findings in recent years in various social contexts. However, we show that a unique feature in the Israeli case is that social advantage is also related to ethnicity and religiosity. We elaborate on the role of these correlates in shaping cultural consumption patterns in Israel.

Regarding class and status, the theoretical framework that draws on Weber’s distinction between the two dimensions maintains that cultural consumption is mainly associated with status. It is revealing, then, that in Israeli society class affects cultural consumption patterns and status does not (for recent studies on the effect of status versus class on consumption see Poetics volume 35 and citations therein). How can this finding be reconciled with the theoretical hypothesis? Are there certain conditions under which we should expect a particularly strong association between economic resources and consumption practices? We argue that such associations are conditioned by the degree to which cultural policy regulates the cost of cultural consumption. In other words, cultural policy can offset the effect of economic resources.

Cultural policies have historically contributed to cultural distinctions along class lines (Beale 1999). State policy can reduce constraints on cultural consumption in various ways, such as regulating cultural activities, providing cultural services through state cultural institutions, and subsidizing individual cultural products. Cultural policy can thereby attenuate the effect of economic resources. In Israel, public expenditure on culture is decreasing, and is directed mainly toward popular culture (Katz 1999). Moreover, the level of state subsidies to the arts, and particularly to the fine arts, is very low compared with other countries. For example, a 2001 parliamentary report showed that the level of public expenditure on the arts in Israel was only 0.02% of the GDP, compared with 0.14% in Britain, 0.26% in France, and 0.21% in the Netherlands (Fishman and Ben-Dror 2001). We argue that this may explain why, contrary to our theoretical expectation, we find that class rather than status is so central in determining cultural consumption in Israel.Footnote 10

We now turn to the roles of religiosity and ethnicity in shaping cultural consumption in Israel. We show that religiosity is negatively associated with cultural participation, and that among those who participate, the more secular are more omnivorous. To explain this finding we recall that membership of a religious group in Israel sets limits on cultural participation. Religious individuals are expected to strictly follow a particular lifestyle and the ultra-orthodox even more so. Strong social control and peer group pressure are at work in their communities, which are reinforced because of residential segregation. This is not to say that religious individuals in Israeli society do not participate in cultural activities: all we can say is that they are less likely to participate in those activities that we measured, and that are considered part of desirable secular cultural capital. Had we measured activities such as attending lectures by Rabbis, participating in study groups, or joining in religious rites and rituals, our results may have been different.

Another traditional cleavage in Israel is ethnicity. Ashkenazi and second-generation Israeli Jews are more likely to be culturally active, and among the active population they are more likely to be omnivorous than non-highbrow actives. This resonates with the social position of the Ashkenazi group, considered the super-ordinate Jewish ethnic group in Israel. The characteristics of the dominant cultural capital in Israel are similar to those in other Western countries. In Israel, however, another meaningful dimension of national cultural capital exists. This encompasses cultural competencies that provide natural membership in the Israeli collective. Core elements of Israeli Hebrew culture were constituted in the formative years of settlement in Israel (1920–1950) by Ashkenazi Jews. These elements are perceived as reflecting a Western secular standpoint that excludes and weakens other (e.g., Mizrachi and Religious) cultural alternatives.

Traditionally, research on the stratification of cultural consumption and taste tended to emphasize the role of class in shaping cultural behavior and preferences (Bourdieu 1984). Our study highlights the importance of additional axes of social advantage that stand out in the Israeli context, but are surely not unique to Israeli society. Indeed their importance is likely to increase in Western societies that have experienced high rates of immigration, which changes their ethnic makeup and their religion composition. Acknowledging ethnicity and religiosity as important dimensions in the generation of cultural advantage should prompt a multifaceted conceptual framework for the study of cultural stratification in multicultural societies.