H
O
?
C
E
Omar Lizardo a and
Sara Skiles
University of Notre Dame, Department of Sociology, 810 Flanner Hall, Notre Dame, IN, 46556
Tel: (574) 631-1855 Fax: (574) 631-9238
Email: olizardo@nd.edu
Last Revised: July 31, 2007
Words: 9,383 a Corresponding Author.

1
Abstract
To date, Peterson and Kern’s (1996) “highbrow omnivorousness” hypothesis has been examined mainly for the case of musical taste. In this paper we attempt to extend this framework to a relatively unexplored cultural domain, that of television consumption. Using data from the 2001 Eurobarometer we hypothesize that, in accordance with the highbrow omnivorousness thesis, highbrows will be more likely to consume a wide variety of other forms of popular culture, namely television programming. The results fail to unambiguously confirm the highbrow omnivorousness hypothesis: in some EU countries, highbrows consume a wider variety of television programming than non-highbrows, in other countries, highbrows are indistinguishable from non-highbrows while in a third group of countries, highbrows are snobbier than non-highbrows in their television consumption choices. We attempt to explain this cross-national heterogeneity in the highbrow/non-highbrow difference in television consumption using DiMaggio’s (2002[1977]) organizational theory of culture production and Crane’s (1993) classification of culture production domains. In our
“contingent highbrow omnivorousness” framework, we propose that in cultural production domains most likely to be structured in ways that promote “mass cultural” styles of production (such as commercial television), highbrow snobbery rather than omnivorousness will be the norm. In culture production fields that are relatively less dominated by market logics on the other hand, the highbrow snobbery effect will be weaker. Classifying countries by the degree of market orientation of the television production field yields results that are consistent with this hypothesis.

2
1 I
In a much cited article, Peterson and Kern (1996) put forward and subject to empirical test their “highbrow omnivorousness” hypothesis. They challenge the standard picture of a radical incompatibility between the consumption of the most “legitimate” and
“consecrated” forms of culture (such classical music, the plastic arts) and the consumption of either mass produced popular culture or other “lowbrow” and “middlebrow” cultural forms. Instead, they propose that persons who choose traditionally elite cultural forms will also be more likely to enjoy as wide a variety of other popular and folk arts as other individuals.
Using data from the 1982 and 1992 waves of the Survey for Public Participation in the Arts
(fielded in the United States), Peterson and Kern (1996) find support for this hypothesis.
Classifying musical genres as either “lowbrow” or “middlebrow” and respondents as either
“highbrow” or “other” they find few or no statistically significant differences between the average number of the former two types of musical genres chosen by highbrows
(operationalized as those who picked either opera or classical music as their favorite music and who reported consuming both of those genres) and the rest of the respondents.
In addition, they find that these differences are much smaller for the more recent period, with highbrows in 1992 choosing on average a higher number of middlebrow genres than other respondents. They conclude (1996: 902) that “....these findings suggest that in 1992 highbrows…have become more omnivorous than others” and that (1996: 903-904) “[t]aken together, the findings…support the assertion that omnivorousness is replacing snobbishness

3 among Americans of highbrow status.” Peterson and Kern’s results served to confirm similar empirical findings examined in earlier research by Peterson and Simkus (1993) and Peterson
(1992) at the aggregate level of occupational groups, as well as the theoretical predictions of
DiMaggio (1987).
1
The most recent work in this field has provided a more nuanced picture of omnivorous taste.
As Warde, Wright and Modesto-Cal (2007) explain, the concept of the omnivore is being replaced by the idea of multiple variations of omnivores. Recent analysis of British, Dutch and American data has revealed that highbrow taste is not a necessary component of omnivorousness (Warde, Wright, Modesto-Cal 2007; van Eijck and Knulst 2005; Peterson
2005a). That is, the definition of omnivore has been widened to incorporate individuals who consume a wide range of popular culture products but few highbrow goods, if any. Broadly, this shift is a result of younger cohorts entering the market and exhibiting different taste patterns than members of previous cohorts, perhaps due to differences in socialization (van
Eijck and Knults 2005).
In this paper we broaden the scope of research on the highbrow omnivorousness phenomenon in two primary ways: first, we further extend Peterson and Kern’s original hypothesis into a cross-national comparative context (Katz-Gerro 2002; Peterson 2005a), by using survey data from 15 EU countries. Previous cross-national work has compared consumption patterns within the countries of the United Kingdom (Gayo-Cal, Savage and
Warde 2006), between Britain and Switzerland (Lambert et al 2005), and between Italy,
1 Various versions of the omnivorousness thesis—mostly related to the higher propensity of the well educated to consume all types of cultural fare—have been tested and broadly confirmed in a wide variety of different national and cultural contexts. Peterson (2005a) and Peterson and Anand (2004) provide the most recent surveys of this growing literature.

4
Israel, West Germany, Sweden and the U.S. (Katz-Gerro 2002, 2006) but none have incorporated data from as many as 15 countries. Second, we continue the recent research trend of investigating the highbrow omnivorousness effect, which has for the most part been observed in the realm of musical consumption, in other cultural domains (Bennett
2006; Bukodi 2007; Fisher and Preece 2003; L
ó pez-Sintas and Garc
í a-
Á lvarez 2006; van Rees et al 1999). We do this by attempting to ascertain whether respondents who exhibit a highbrow taste pattern are more omnivorous in their consumption of a wide variety of
popular television programming, thus further extending the range of applicability of the highbrow omnivorousness hypothesis to the realm of media culture production (Crane 1993).
In this manner we join other sociology of taste scholars in attempting to advance Peterson’s
(2005a: 267) suggestion to test “…the omnivore idea across the full range of style choices.”
The contrast between television and music is of both substantive and theoretical interest.
We draw on Crane’s (1993) differentiation between the three different “domains” of culture production in contemporary (post)industrial societies, what she refers to as the “core”
(national and global culture industries such as television and movies), “peripheral” (niche marketing culture industries such as music and publishing) and the “urban” (small scale culture production in “culture-worlds” and “art-worlds” populated by small bands of independent producers and non-profit delivery and dissemination organizations) domains.
Crane (1993) offers this scheme in order to transcend the conceptual limitations of the usual elite/popular distinction in the sociology of culture.
Taking into account Crane’s typology, one important limitation of the usual focus on music and literature consumption in the recent literature (i.e. Bryson 1996; Bukoki 2007;

5
Coulangeon and Lenel 2007; Katz-Gerro, Raz and Yaish 2007; Sonnett 2004; Torsche 2007; van Eijck 2001; van Rees et al 1999; Wright 2006) is that most of what is assumed to be known regarding the omnivore thesis has been based on patterns of consumption related to
Crane’s “peripheral” domain (music and book publishing). Less is known however about differences in the patterns of consumption of highbrows versus others in Crane’s “core” domain (and much less in what she refers to as the “urban” domain), even though arguably the core domain is where most of what Alexander (2003) refers to as the “popular arts” that reaches entire national populations is produced and disseminated. Scholars working in multiple Western countries have found various differences in viewing behavior and genre preferences in television audiences, including distinctions by age, gender, social status and education attainment (Bennett 2006; Bennett, Emmison and Frow 1999; Bihagen and Katz-
Gerro 2000; Kraaykamp et al 2007; Kuipers 2006; L
ó pez-Sintas and Garc
í a-
Á lvarez 2006; van
Eijck and van Rees 2000). In addition, Blewitt (1993) offers a Bourdieuian explanation for differences in audience composition for Art House and Hollywood films, and Chan and
Goldthorpe (2005) find a stratified cinema attendance pattern. By focusing on television, one of the primary aims and contributions of this study is to further extend research on omnivorousness to this crucial site of culture production and everyday consumption (Fiske
1987). This comparison is of particular theoretical interest since the appeal of the omnivore thesis has been primarily sustained by the fact that its empirical implications are so largely at odds with traditional “mass culture” approaches (DiMaggio 1987). Therefore, we find that further empirical tests of patterns of consumption in realms of cultural choice that would be expected to be closest to this category of cultural goods constitute a significant contribution to the cultural consumption literature (DiMaggio 2002[1977], 1991).

6
2 D
H
In this study we use micro-level data of the culture consumption habits of EU citizens taken from the August-September 2001 Eurobarometer (Christensen 2003). The Eurobarometer is designed to provide regular monitoring of the social and political attitudes of the EU population. Since the early seventies representative national samples in all European Union
(formerly European Community) member countries have been simultaneously interviewed in each spring and each autumn. More recent versions of the Eurobarometer (starting in the autumn of 1990) have included supplementary surveys on special issues of topical interest.
The 2001 Autumn Eurobarometer (N=16,200) included such a special “module” related to participation in a wide variety of cultural activities, from mass-media (radio and television) to music and the arts. The Eurobarometer includes stratified probability samples of citizens of the EU aged 15 and over residing in the 15 EU countries: Austria, Belgium, Denmark, Finland,
France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain,
Sweden, and the United Kingdom.
2
The main empirical expectation in our study is the “highbrow omnivorousness” hypothesis as applied to television consumption, and as a replication of Peterson and Kern’s (1996) study, to musical consumption. Thus, if the highbrow omnivorousness thesis is applicable to music consumption in Europe in the very same way as Peterson and Kern (1996) demonstrated that it was applicable for the U.S., then we should expect that:
2 Samples sizes for individual countries range from a low of 609 for Luxembourg to a high 1,093 for
Austria. For most of the other countries the sample size hovers around 1000. The combined sample for East and West Germany consists of about 2,000 respondents. In order to keep a balanced number of respondents across all countries we exclude the East German sample from the analysis. This leaves us with about 1000 respondents for (West) Germany.

7
(1) Highbrow omnivorousness hypothesis (music): across all EU countries, highbrow respondents are either more or equally likely to consume non-elite musical genres as other respondents.
Extending this same logic to the consumption of popular television we should find that:
(2) Highbrow omnivorousness hypothesis (television): across all EU countries, highbrow respondents are either more or equally likely to consume different types of popular television programming as other respondents.
3 V
3.1
M USIC
Each respondent in the Eurobarometer survey was asked the following question: What kind of music do you listen to? Responses were organized into 11 broad musical categories: 1) classical music, 2) opera or operetta, 3) rock and roll and pop rock, 4) heavy metal or hard rock, 5) easy listening, 6) electronic dance music or “house,” 7) techno or ambient, 8) rap or hip hop, 9), folk or traditional music (including American country music) and 10) world music.
The respondent is counted as having consumed the musical form if she reports having listened to a recorded performance of the genre at least once in the past month. The survey also asked which type of musical concerts respondents had attended in the previous month, using the same 11 categories as listed above. For each genre we assigned a one to each respondent if he or she reported having listened to or having attended a concert featuring that genre. We then added all of the responses to create an omnivore consumption scale.

8
Across all 15 countries, respondents report consuming 1.90 non-elite music genres on average.
3
We follow Peterson and Kern (1996), Peterson (2005b) and Peterson and Rossman (2007) in using musical taste items to differentiate between highbrow and non-highbrow respondents.
This is consistent with Bourdieu’s (1984: 18-19) contention that, “there is no more
‘classificatory’ practice than concert going or playing a ‘noble’ instrument…because the flaunting of ‘musical culture’ is not a cultural display like others… [Music] represents the most radical and most absolute form of the negation of the world, and especially the social world, which the bourgeois ethos tends to demand of all forms of art.”
However, an important methodological issue emerges in the process of defining highbrow status, since the analyst must take into account changing patterns of legitimation of different musical forms. In a certain sense—to use Williams’ (1973) terms—we have to differentiate between currently dominant, residual and emergent patterns of highbrow taste.
For instance, while considering Opera an indicator of highbrow taste—as done by Peterson and Kern (1996)—makes sense in the case of the United States, where Opera performances, aided by their European exoticism, were actively institutionalized as part of upper-middle class leisure practices early in the 20 th century (DiMaggio 1993). However, in many European countries, Opera is part of the popular and folk arts realms (Storey 2003). We must also take into consideration the relatively recent rise of jazz appreciation to the ranks of the highbrow canon (Peterson 1972). Thus, the cultural mobility of different genres means that what
3 There is suggestive cross-national variation around the grand mean, with the average number of nonhighbrow genres consumed ranging from a low of 1.44 in Italy to a high of 2.34 in Finland.

9 counts as highbrow may be historically specific (Peterson 1992, 1997, 2002; DiMaggio 1993;
Levine 1984).
To that end, we take a data-driven approach to the operationalization of highbrow status.
Table 1 show the sociodemographic characteristics of respondents arranged by the eight
possible response patterns (Gilula and Haberman 2001) of highbrow consumption from the all the possible cross-tabulations of binary indicators of classical, opera and jazz music appreciation (defined as having listened to that genre at least once in the past month). Two patterns appear to emerge as unambiguously highbrow in the dominant sense, as measured by the disproportionate likelihood of respondents displaying these patterns to be college educated, urban, consumers of the plastic arts and of belonging to a managerial or professional occupation: the classical-jazz pattern (6) and the classical-opera-jazz pattern
(8). Two other patterns, classical only (5) and the classical-opera (7) pattern appear as somewhat dominant as measured by the education indicator (in particular the last one) but not by the occupation indicator. Further, as given by the average age of the respondent for the classical only pattern, this appears to be giving way to the first two patterns (it seems to have become a residual pattern of highbrow taste). A taste for jazz only (2) on the other hand seems at this stage to be an emergent and not a dominant pattern. A taste for opera only, does not appear to be a highbrow pattern at all as it is not positively associated with sociodemographic markers of status and privilege. This is consistent with the ambiguous status of opera appreciation in the European context. We thus consider a respondent to be of highbrow status if the respondent displays consumption patterns 6, 7 or 8 as shown in table 1, the patterns that appear to be most unambiguously “dominant.” We consider all

10 other respondents as non-highbrow. Under this definition, about 13.87 of respondents in the
Eurobarometer survey fall under the highbrow category.
4
3.2
T
ELEVISION
Respondents were also asked to report the types of television programs that they customarily watched. We count a respondent as liking a given type of television offering if he or she reports having watched a program in that category at least once in the past month.
There are eight categories that we consider popular television: 1) soap operas/TV drama series, 2) music programs, 3) sport programs, 4) films and movies, 5) children’s programs, 6) talk shows, 7) home shopping programs, 8) Entertainment programs (a category that includes game shows, cooking shows, reality television, and other similar fare). Factor analysis results—not shown—suggested that neither television news nor documentary watching could be considered popular television (as they loaded in a separate latent factor from the other items and had a high correlation with highbrow status as here defined), so we exclude them from our measure of popular television omnivorousness. Across all 15 EU countries respondents chose an average of 3.18 types of popular television programs.
5
4 R
4.1
M USICAL CHOICES
4 Including pattern (5) into our operationalization of highbrow does not substantively change any of the results reported below, but it does increase the proportion of the population that are considered highbrow considerably to 27.01%. We therefore prefer to stick to our more restricted definition of highbrow status.
5 Once again there is important cross-national variation around this overall mean. This ranges from a low of 2.81 television genres chosen in Austria to a high of 3.80 in Finland, the most cultural active country in the Eurobarometer.

11
Table 2 shows the mean differences across all 15 EU countries in the average number of nonelite musical genres chosen by highbrows and others, and Table 3 the average number of popular types of television programming reportedly consumed by these same two groups.
The last two columns of each of the tables report the value of the F statistic and the probability associated with a test of the hypothesis that the difference between highbrows and others is statistically significant from zero for that country. Looking at table 2, we find strong support for Peterson and Kern’s (1996) highbrow omnivorousness hypothesis in the
European context for the case of music consumption. In 2001, for 12 out of 15 (80%)
European societies, highbrows are either statistically more likely to choose a higher number of non-elite musical genres than non-highbrows, or, in the case of Denmark and Sweden, choose as many genres as non-highbrows. The only exception to this pattern are the
Netherlands, Germany and Austria, where highbrows appear to be snobbier than nonhighbrows (p<0.05).
Why are Germany and Austria distinctive? We venture to speculate that as the primary focus of the institutionalization of the classical aesthetic “ideology” developed in Vienna in the 18 th and 19 th centuries (DeNora 1991), the Austro-German region might also be displaying a specific pattern of entrenched resistance to the “declassification” processes that are part and parcel of the decline of the traditional “snob” aesthetic and its replacement by omnivore standards (DiMaggio 1987; Peterson 1992, 1997, 2005a).
The Netherlands, however, still appears as an anomalous case. It is important to note however, that when highbrow status is defined in the most restrictive manner possible
(using the two patterns that are most clearly dominant in table 1: patterns 6 and patterns 8

12 and excluding pattern 7), that the Netherlands reverts from a negative difference between highbrows and others, to a positive difference (p<0.05) a pattern consistent with the highbrow omnivorousness thesis. In a similar way, when using this definition of highbrow status, the difference between highbrows and others in Germany and Austria becomes statistically indistinguishable from zero, a pattern also consistent with the Peterson thesis.
Thus, it appears that the bulk of the “snob” effect in these three countries is being disproportionately carried by those highbrows who combine classical and opera but exclude jazz. The fact that this group is more exclusivist is not surprising, since, as shown in table 1, it is disproportionately composed of the “declining fraction” of older highbrows (Bourdieu
1984: 346-351), who are less likely to be employed in a “new class” managerial professional occupation and who thus appear to form a conservative “rear guard” (van Eijck and
Bargemann 2004). The rising, “new-fangled” highbrows—or those younger members of the upper middle class who put Jazz into the elite mix—are less likely to be exclusivist in relation to non-elite genres.
Going back to Table 2, we find that in most European countries the highbrow/other
difference in omnivore consumption hovers around an advantage of little less than one genre (0.82) on average. Nevertheless there are suggestive cross-national variations around this central tendency. For instance, in Greece and Portugal, the differences between highbrows and others are much more dramatic than this, with highbrows consuming about two more genres than non-highbrows on average (a difference of 1.91 and 2.24 genres, respectively).

13
How to explain this variation? It turns out that the magnitude of the highbrow non-highbrow advantage appears to be connected with cross-national differences in the average age of the highbrow culture group. Thus, while highbrows tend to be older than non-highbrows across most European societies, Greece and Portugal are the only two countries in which this pattern is reversed, with highbrows being younger than non-highbrows on average.
Furthermore, the countries in which highbrows are the oldest, are also the countries in which the differences between highbrows and others are smallest (or even negative). There is a zero-order correlation of -0.83 between the age difference between highbrows and others and the omnivore consumption difference between these same two groups across all fifteen countries.
This suggests, in accordance with van Eijck’s (1999) proposal, that highbrow omnivorousness has a lot to do with generational differences between different segments of the upper middle class. Countries in which the “new class” of highbrow consumers is comparatively younger are also the ones where the starkest differences in highbrow versus non-highbrow consumption can be observed. This is presumably because younger highbrows are much less likely to forgo the consumption of popular and folk cultural forms even as they come to adopt the traditionally cultural habits of the upper-middle class (Peterson 2005a). These differences appear to be exacerbated in some Southern European societies (with a legacy of ethnic and linguistic heterogeneity and Agrarian provinces outside the major metropolises) in which the “univore” pattern of consumption among the non-highbrow culture class is more likely to be the rule.
4.2
T ELEVISION PROGRAMMING CHOICES

14
Hypothesis 2 suggests that a similar highbrow omnivorousness pattern discussed by
Peterson and Kern (1996) and which obtains in the present case for music consumption
should also be observed for television consumption. Table 3 shows the results. In contrast
to the mostly uniform and systematic pattern of highbrow advantage in omnivore consumption over non-highbrows, the results for television appear to be much more ambiguous.
Highbrows tend to display a snobbish consumption pattern (consuming a lower variety of television programming on average) in Denmark, France, Spain, The U.K. and Finland
(p<0.05) and to a somewhat weaker extent in Austria (p<0.10), a series of results inconsistent with the omnivore thesis. In Portugal on the other hand, highbrows consume a larger number of television program types than non-highbrows, while in Belgium, Germany, Greece,
Italy, Ireland, Luxembourg, the Netherlands and Sweden, highbrows appear to be statistically indistinguishable from non-highbrows, a pattern of results that is consistent with the omnivore thesis. There is variation among this last group of countries however, with some countries showing a highbrow advantage pattern and others a highbrow snobbishness pattern. Thus, it appears that Peterson and Kern’s (1996) highbrow omnivorousness generalization cannot be unambiguously applicable to a core domain such as television consumption.
6
How can we explain this cross-national heterogeneity in the highbrow-non-highbrow difference in television consumption? It is possible that the apparent lack of clear
6 Using the more restrictive definition of highbrow status that removes the older rearguard (pattern 7 in table 1) from the operationalization, produces results substantively identical to those reported in table 3.

15 applicability of the highbrow omnivorousness hypothesis in the case of television consumption might have something to with the fact that the mechanisms governing the
“style choices” of cultural products that belong to Crane’s (1993) “core domain” are different from those that pertain to a peripheral domain such as music. This suggests that the highbrow omnivorousness effect can be directly linked to the larger organizational structure of the realm of culture production from which choices are made. 7
We can hypothesize that the degree of clear-cut applicability of the highbrow omnivorousness hypothesis will increase the closer a particular cultural real is to Crane’s
“peripheral domain” (as in music) and will decrease—with individuals reverting to a highbrow snobbishness pattern in some cases—the closer a particular set of cultural goods are to the “core domain” of mass culture production. “Reversals” to the highbrow snobbishness pattern, may then be tied to the extent to which the core domain in question fits the ideal type of a “mass culture industry” (Adorno 2001). This means that the observed variation in the highbrow/non-highbrow difference in the case of television consumption choices might be related to cross-national differences in the way in which the television production field is organized.
Is there any way to get some empirical leverage on this hypothesis? In the following, we take advantage of cross-national differences in the institutional and organizational structure of television production and dissemination to examine this possibility.
7 There is no significant association between the highbrow/non-highbrow cross-national difference in musical choices and television programming choices (r=0.21, p=0.44), suggesting that they are driven by distinct mechanisms.

16
5 A “
”
In developing our “contingent highbrow omnivorousness” framework we turn to DiMaggio’s
(2002[1977]) classification of cultural industry systems along a dimension that separates those culture production systems that are organized in accordance to a production logic that stresses reproducibility, sameness and a market orientation and those which stress uniqueness, producer autonomy and a non-profit orientation. DiMaggio explains that the mass culture industry system has been criticized for the homogeneity of its goods and for favoring predictability over innovation in its appeal to the lowest common denominator of taste in effort to attract the maximum number of consumers (Adorno 2001). This perspective also faults consumers for their uncritical acceptance of the products offered in the mass market. DiMaggio (2002[1977]) contends that although such cultural uniformity is possible, other more diverse and creative outcomes are just as likely, and that the extent to which innovation and creativity are mobilized is a function of market mediation, industry structure and the organization of work.
Using Peterson and Berger’s model (1975), DiMaggio explains that the combination of high market concentration and centralized industry structure characteristic of “mass” culture industries (such as television) results in low innovation and diversity. He contrasts such an industry structure with “class” and “pluralistic” systems in which lower market concentration and decentralized industry structure result in greater innovation and diversity
[Dowd (2004) explains that decentralization mediates the negative effects of concentration on diversity). This culture production model assumes that consumers have a high demand for a diversity of options (in contrast to the mass culture critique) but that profit-oriented

17 production companies prefer to focus on a narrow range of cultural goods to minimize risk involved with promoting untested products.
This model informs our discussion of highbrow consumption, indicating that omnivorousness is more likely in a market structure characterized by decentralized industry organization, producer autonomy and product diversity (such as music) rather than low innovation (such as television). In contrast to DiMaggio (2002[1977]) who traces differences in industry structure across different culture production domains (television, music, publishing, etc.) we capitalize on the fact that an analogous degree of variation along some of DiMaggio’s dimensions can be observed cross-nationally across all fifteen EU countries within the realm of Television programming production. In particular we focus on variation along the degree to which a particular system of television production is market oriented versus state
dominated. We hypothesize that market-oriented systems are closer to DiMaggio’s “mass culture” pole and state-dominated systems are relatively closer to his more diverse “class culture” pole.
(3) Contingent omnivorousness hypothesis: In EU countries in which television programming is less oriented toward the market and more toward the non-profit arena highbrow respondents will be more or equally likely to consume different types of popular television programming as non-highbrows. In market oriented regimes, on the other hand, highbrows will display snobbish patterns of consumption in comparison to
non-highbrows.
5.1
M EASURE OF TELEVISION INDUSTRY STRUCTURE

18
For each country we use the average rank of each country on three macro-level indicators of the degree of market penetration in the television industry as our measure of the market orientation of the television production field: (1) the logged ratio of the per capita number of for-profit television stations to the number of state-owned television stations; (2) the proportion of television stations that are private; (3) an estimate of the amount of Broadcast
Television Economic Infrastructure (BTEI).
8 Each of these variables was converted into decile ranks and the average rank—with larger absolute values of the rank indicating higher levels of market orientation—for each country was computed for the three indicators. This average rank is our measure of the overall market orientation of the Television production field. We should expect, if hypothesis 3 is on the right track, that the highbrow-other difference is much starker in market-oriented television production regimes (closer to
Crane’s ideal type of “core” popular culture production and DiMaggio’s “mass culture” system), and much smaller (or even null) in non-profit oriented regimes.
5.2
R ESULTS
Figure 1 shows a scatter plot with the average market-orientation rank score on the X-axis and the highbrow-other popular television consumption difference in the Y-axis, for each one of the 15 EU countries included in the 2001 Eurobarometer. The results are broadly consistent with our contingent highbrow omnivorousness hypothesis. The correlation between these two variables across the fifteen EU countries is a robust -0.74. Save for for the Netherlands and Greece (and to a lesser extent Italy), the only countries with a highbrow/non-highbrow difference close to zero but with relatively market-oriented
8 The television station per capita data were obtained from Vissol (2006: 33, table 1) and the BTEI and proportion private data from Dupagne and Waterman (1998: 214, table 1).

19 television production regimes, all other European societies fall neatly around the implied regression line with a steep negative slope. Portugal displays higher levels of highbrow omnivorousness than expected (mostly due to the distinctive age composition of the highbrow group in that country), but it is toward the lowest observed end of market orientation.
Thus, it appears as shown in the figure that the relative snobbishness of highbrows when it comes to television consumption is highly dependent upon the structure and institutional organization of the national television production field. In countries in which Television programming has been largely penetrated by the market (i.e. Spain, France, and Finland) highbrow snobbery in relation to television is relatively stronger. In countries in which television production is relatively dominated by the state (Portugal, Ireland, Belgium), highbrow snobbishness is much weaker, and in the case of Portugal, non-existent, as highbrows appear to be omnivores even in the realm of television consumption. This certainly does not seem to be the only explanation (French highbrows, for instance, display higher levels of snobbish avoidance of television than expected given their levels of market orientation), but it does appear to capture a good part of the systematic influences that appear to be driving the cross-national differences.
5.2.1
R ANDOM -E FFECTS M ODELS
Table 4 provides a more rigorous test of the contingent highbrow omnivorousness
hypothesis. The table presents the coefficient estimates of a series of random-effects regression models with the omnivore television score as the dependent variable. Randomeffects models take into account cross-national heterogeneity in the pooled sample by

20 including an additional error term (μ j
, j=1, 2, 3…15) in the regression equation which in contrast to the usual regression disturbance—which is different for each individual—is allowed to vary across countries but is the same within countries. Thus the additional random effect can be taken as a measure of all cross-national heterogeneity—as measured by the estimated variance of the country-specific error term (σ[μ j
])—that is not accounted for by the individual and country-level variables included in the model.
Model 1 is a baseline model which essentially reproduces the results shown in Table 3: on
average across all 15 Eurobarometer countries, highbrows consume about -0.2 less television genres than non-highbrows, even after taking unmeasured cross-national variation into account. This highbrow snobbishness effect concerning television programming choices contrasts sharply with the observed pattern of relative omnivorousness in relation to musical consumption. However, we know from figure 1 that there is important cross-national variation around this overall mean difference (with some countries even displaying the opposite or a null effect) and that a good portion of this unmeasured variation revolves around the market-structure of television production.
To explicitly account for this source of cross-national heterogeneity, in model 2, we introduce the (centered at its grand mean) country-level measure of television market orientation, and an interaction between that measure and the highbrow dummy variable. The resulting interaction effect can be thought of as an estimate of how the coefficient corresponding to highbrow status varies cross-nationally in reaction to increasing levels of television market orientation. Consistent with the contingent highbrow omnivorousness hypothesis, the interaction effect is negative and statistically significant (t=-3.94), suggesting that highbrows

21 embedded in market oriented television production regimes—assumed to be more likely to produce “mass cultural” forms of television (DiMaggio 2002[1977])—are even snobbier than the cross-national average (about 45% more so). The insignificant main effect of marketorientation (t=0.25) suggests that the structure of the television production field has no impact on the relative breadth of choices of non-highbrows, in contrast to its depressing effect among highbrows.
In model 3, we introduce some standard sociodemographic controls. We focus on those individual level variables that are most likely to be associated with heavy television usage, in particular age, gender, education, and certain forms of labor market status such as being retired or being a full-time homemaker.
9 After controlling for the relevant sociodemographic characteristics, we find that—as given by the main effect for this variable—the effect of highbrow status is greatly reduced
,
and is not longer statistically significant (t=-0.36). This suggests that in EU countries that are at about at the average or lower in terms of television market orientation, highbrows are no different from others in their levels of television consumption after controlling for the relevant socio-demographic indicators. This result is consistent with Peterson’s highbrow omnivorousness thesis and the plot shown in figure 1.
However, the negative mediating effect of market orientation on highbrow snobbery remains virtually unchanged between the two models (t=-3.76). This means that in countries in which the television market is highly oriented toward the for-profit sector, highbrows display snobbish patterns of television programming choices in relation to non-highbrows.
The fact that the basic result remains even after controlling for individual-level
9 Means and standard deviations of the sociodemographic variables are available on request.

22 heterogeneity, suggests that this effect has little to do with individual differences between highbrows and non-highbrows (or cross-national differences in population composition) related to age, education, gender, and the other variables included in model 3, but that it is a macro-level mediating effect of the industry structure of television production on the behavior of highbrow respondents.
6 D
6.1
S
UMMARY OF THE RESULTS
In this paper we have attempted to extend the notion of highbrow omnivorousness to television consumption, . This is a culture production domain that is sufficiently different in theoretically relevant ways (Crane 1993) to produce an illuminating contrast to music consumption, the cultural domain that has been the focus of the bulk of research on omnivore consumption to date. The results show that Peterson and Kern’s (1996) highbrow omnivorousness thesis cannot be unproblematically extended to this realm.
When it comes to television consumption, there is wide cross-national variation in the highbrow/non-highbrow difference. In some countries, highbrows either consume a wider range of television programming than non-highbrows or have statistically indistinguishable patterns of cultural choice, but in other countries, highbrows tend to have much more restricted (“snobbish”) patterns of television consumption than non-highbrows. This is in spite of the fact that highbrows tend to almost uniformly conform to the Peterson-Kern
(1996) notion of omnivorousness in the realm of musical consumption across the majority of

23 countries in this same data set and universally across all countries when highbrow status is defined in a more exacting manner.
We attempt to explain these apparently inconsistent results by integrating individual-level research on cultural omnivorousness with a more macro-structural “production of culture” approach (Peterson 1976, 1979; Peterson and Anand 2004; Crane 1993; DiMaggio 2000) to the determinants of the different menu of cultural choices that highbrow individuals embedded in different national contexts of television culture production might be exposed to. Our contingent highbrow omnivorousness argument is premised on the assumption that highbrow inclusiveness is in large part dependent on the style of production that is dominant for the cultural domain in question. In peripheral culture production domains as argued by
Crane (1993), niche marketing and more “artisanal” styles of cultural good production will tend to predominate, facilitating the creation and diffusion of “class culture” and
“subcultural” goods (see also DiMaggio 2002[1977]), which carry sufficiently aesthetic merit to fall under the all-inclusive “aestheticizing gaze” of the omnivore (Holt 1998), or what
Bourdieu (1984) called the generalized “aesthetic disposition.”
The core culture production domain, on the other hand, is characterized by the routinized and rationalized (in the Weberian sense) processing of highly standardized, nationally and
(increasingly) globally disseminated cultural fare, with cultural creators embedded in large bureaucratic for-profit organizations and—when the logic of the market predominates— increasingly subject to severe restrictions in creative autonomy and independent initiative
(Alexander 2003; Bourdieu 1999; Crane 1993; Hirsch 1972). This is a style of culture production that is very likely to produce “mass cultural” goods (DiMaggio 2002[1977]).

24
We theorize that highbrows are likely to stay away from mass cultural forms (Adorno 2001), and that this explains the differences between their relative omnivorousness in the case of music and the inconsistent results when it comes to television consumption. To attempt to shed empirical light on this last question, we take advantage of the cross-national nature of the data set at hand by looking into the relative variation of the market-orientation of the television production field across societies. Using a macro-level composite measure of market orientation we find results that are consistent with the contingent omnivorousness hypothesis: in countries characterized by a heavily market-penetrated television production field, the highbrow snobbishness effect is much stronger than in countries that in which for profit television stations are less prevalent in relation to the non-profit state dominated sector.
6.2
THEORETICAL IMPLICATIONS
This study advances theory and research on cultural omnivorousness in several ways: first we show that Peterson and Kern’s (1996) and Peterson and Rossman’s (2007) framework for the study of the differences in culture consumption between highbrow and non-highbrow respondents can be extended to other cultural realms beyond music. However, we show that the notion of “highbrow omnivorousness” is by no means universal across cultural
realms, but that it may be highly dependent on the domain in which the majority of the cultural goods associated with that set of cultural genres are produced. Second, drawing on
Crane’s (1993) division between the three primary domains of culture production, we suggest the following generalization: highbrow omnivorousness is increasingly likely to be found for cultural goods that are produced in the peripheral domain (such as music and

25 literature), but is relatively less likely to obtain for cultural goods produced either in the core
domain (national systems of film production and television) or the urban arts domain, which is more likely to be the home of subcultural and other avant garde movements advanced in urban “genre worlds” (Negus 1999: 29). Cultural styles produced in the urban domain may come to form part of the highbrow repertoire once they are taken out of this domain by cultural entrepreneurs and transferred in some form or another into the peripheral domain
(as appeared to have been the case for Jazz [Peterson 1972]). Furthermore, core industries that take on some of the specificity and relative creative autonomy of peripheral domain industries (i.e. independent film and documentaries) may increase the prevalence and intensity of highbrow omnivorousness in those realms.
Third, these findings of differential highbrow consumption in core and periphery domains can be explained, at least in part, by Crane’s earlier work (1992, ch. 4), which updates
Peterson and Berger’s (1975) model to take into account market fluctuations in the 1970s and
1980s. She finds that during this time, core domain conglomerates had the financial resources to maintain market control during turbulent times, meaning that they were not forced to innovate when concentration decreased, but could maintain previous standards.
Such was not the case for periphery domain markets that found it necessary to keep up with emerging trends in order to maintain market control (Lopes 1992 and Dowd 2004 explain this dynamic as a result of decentralization). Therefore, market conditions mandate that production processes of cultural goods produced in periphery domains respond to fluctuations in consumer demand at a greater rate than production processes of core domain products. Thus, our finding that highbrows have broader tastes in music (a periphery

26 domain) than in television (a core domain) can be explained by the fact that their desires are better met in the music market than in the television market.
Fourth, our findings confirm previous work by Bennett (2006) and van Eijck and van Rees
(2000) which show that highbrows are most likely to prefer programming on public as opposed to private television stations, and that the type of television shows most liked by highbrows in both the British and Dutch markets are ‘legitimate’ programs, meaning news, current affairs or documentary formats.
6.3
S UGGESTIONS FOR FUTURE RESEARCH
In essence, we join Peterson (2005a) in calling for further comparative research on the relative applicability and generalizability of the notion of highbrow omnivorousness. In the development of the “production of culture” approach, Peterson (1990) examined six factors associated with cultural production: industry structure, organization structure, market, technology, law and regulation, and occupational career. Our analysis of differential television consumption takes industry structure into account. Further studies may benefit from the inclusion of the other facets as well. However, as Peterson and Anand suggest
(2004:328), these six facets may be interdependent aspects of culture production only in corporate capitalist societies. It is possible that different or fewer influences may be at play in differently-organized society such as France or Sweden, so future comparative research in this area should consider this possibility.
Our results lead us to conclude that highbrows prefer cultural products from the peripheral domain rather than the core domain. We have examined only one industry within the core

27 domain. Further support for our conclusion could be drawn from the analysis of other core industries such as film or national newspapers (Crane 1993). In their analysis of highbrow omnivorousness, Lopez-Sintas and Katz-Gerro note that highbrows “prefer to master a wide range of cultural competencies” (2005: 300). If future research supports the conclusion that highbrows prefer peripheral domain products, the next question to be answered will be what it is about such goods that are useful to highbrows as they strive to enhance their cultural competency. That is, why is U2 more attractive to highbrows than Dallas? One answer to this question might have to do with how certain culture-production industries and products are more likely to serve as bases for the deployment of what Ang (1993[1985]) referred to as the “ideology of mass culture.” Also, we suggest that highbrow omnivorousness is increasingly likely to be found for cultural goods that are produced in the peripheral domain than urban domain, at least until urban goods are incorporated in the peripheral domain. Our contingent highbrow omnivorousness framework would be strengthened with research into consumption patterns of urban domain goods, such as local fairs and performances (Crane 1993).
However, as we have shown here, one important way to gain leverage on many of the questions that remain unanswered might be to go beyond the individualistic bias of previous research on omnivore consumption (something that has no direct connection to the fact that a lot of this research has been done using survey data, as multilevel analytic techniques are increasingly available to many researchers), in order to incorporate the insights of the firstgeneration of researchers on the macro-level properties of culture production regimes. In

28 that way the production of culture approach can be joined with more micro-level concerns with the “auto-production” of culture through consumption (Peterson 2000).
One important mesolevel perspective that remains untapped in studies of omnivore consumption is Bourdieu’s (1983, 1999: 39-44) field theory of culture production. There is a clear line of connection between some of the issues touched on in this paper and some foundational concepts in field theory (Benson 1999). For instance the notion of marketorientation that we have exploited has an obvious analogue in Bourdieu’s (1983) idea of
“heteronomy” in a culture production field. Thus, another way of drawing the contrast between Crane’s peripheral and core domain might be to locate them in Bourdieu’s
“autonomy-heteronomy” dimension. This would imply that highbrow omnivorousness in the consumption of a given set of cultural goods increases as the relative autonomy of the field of cultural production in question increases, and it will decrease, sometimes resulting in the opposite pattern, as fields come to be colonized by the market. Future research should concentrate on exploiting this theoretical difference across culture production fields in examining the behavior of highbrows across these contexts. We hope that the research reported in this paper represents one step in that direction.

29
REFERENCES
Adorno, Theodor. 2001. The Culture Industry. New York: Routledge.
Alexander, Victoria. 2003. Sociology of the Arts: Exploring Fine and Popular Forms. Malden:
Blackwell.
Ang, Ien. 1993[1985]. “Dallas and the Ideology of Mass Culture.” Pp. 403-420 in The Cultural
Studies Reader. Edited by Simon During. New York: Routledge.
Bennett, Tony. 2006. “Distinction on the Box: Cultural Capital and the Social Space of
Broadcasting.” Cultural Trends 15:193-212.
Bennett, Tony, Michael Emmison and John Frow. 1999. Accounting for Tastes: Australian
Everyday Cultures. ?: Cambridge University Press.
Benson, Rodney. 1999. “Field Theory in Comparative Context: A New Paradigm for Media
Studies.” Theory and Society 28:463-498.
Bihagen, Erik and Tally Katz-Gerro. 2000. “Culture Consumption in Sweden: The Stability of
Gender Differences.” Poetics 27:327-49.
Blewitt, John. 1993. “Film, Ideology and Bourdieu’s Critique of Public Taste.” British Journal
of Aesthetics 33:367-72.
Bourdieu, Pierre. 1983. “The Field of Cultural Production, or: The Economic World
Reversed.” Poetics, 12:311-356.
———. 1984. Distinction. Cambridge, MA: Harvard University Press.

30
______. 1999. On Television. New York: New Press.
Bukodi, Erzsébet. 2007. “Social Stratification and Cultural Consumption in Hungary: Book
Readership.” Poetics 35:112-131.
Bryson, Bethany, 1996. “‘Anything but heavy metal’: symbolic exclusion and musical dislikes.”
American Sociological Review 61:884–900.
Christensen, Thomas. 2003. “Eurobarometer 56.0: information and communication technologies, financial services, and cultural activities, august-September 2001 [computer file].” 2nd ICPSR version. Brussels: European Opinion Research Group EEIG [producer], 2001.
Cologne, Germany: Zentralarchiv fur Empirische Sozialforschung/Ann Arbor, MI: Inter-
University Consortium For Political And Social Research [distributors].
Chan, Tak Wing and John H. Goldthorpe. 2005. “The Social Stratification of Theatre, Dance and Cinema Attendance.” Cultural Trends 14:193-212.
Coulangeon, Philippe and Yammick Lemel. 2007. “Is ‘Distinction’ Really Outdated?
Questioning the Meaning of the Omnivorization of Musical Taste in Contemporary France.”
Poetics 35:93-111.
Crane, Diana. 1992. The Production of Culture: Media and the Urban Arts. Newbury Park, CA:
Sage.
———1993. “High Culture versus Popular Culture Revisited: Toward a Reconceptualization of Recorded Cultures.” Pp. 58-74 in Cultivating Differences: Symbolic Boundaries and the

31
Making of Inequality, edited by Michèle Lamont and Marcel Fournier. Chicago, IL: Chicago
University Press.
DeNora, Tia. 1991. “Musical Patronage and Social Change in Beethoven Vienna.” American
Journal of Sociology 97:310-346.
DiMaggio, Paul. 1977. “Market Structure, the Creative Process, and Popular Culture: Toward an Organizational Reinterpretation of Mass-Culture Theory.” Pp 151-162 in Cultural Sociology, edited by L. Spillman. Malden, MA: Blackwell.
———. 1987. “Classification in art.” American Sociological Review 52:440–455.
———. 1991. “Social Structure, Institutions, and Cultural Goods: The Case of the U.S.” Pp
133-155 in Social Theory for a Changing Society, edited by Pierre Bourdieu and James Coleman.
Boulder, CO: Westview Press.
______. 1993. “Cultural Boundaries and Structural Change: The Extension of the Highbrow
Model to Theater, Opera and Dance 1900-1940.” Pp. 21-57 in Cultivating Differences:
Symbolic Boundaries and the Making of Inequality, edited by Michèle Lamont and Marcel
Fournier. Chicago, IL: Chicago University Press.
———. 2000. “The Production of Scientific Change: Richard Peterson and the Institutional
Turn in Cultural Sociology.” Poetics 28:107-136.
Dowd, Timothy J. 2004. “Concentration and Diversity Revisited: Production Logics and the
U.S. Mainstream Recording Market, 1940-1990.” Social Forces 82:1411-55.

32
Dupagne, Michel and David Waterman. 1998. “Determinants of U.S. Television Fiction in
Western Europe.” Journal of Electronic and Broadcasting Media 42: 208-220.
Fisher, Timothy C.G. and Stephen B. Preece. 2003. “Evolution, Extinction, or Status Quo?
Canadian Performing Arts Audiences in the 1990s.” Poetics 31:69-86.
Fiske, John. 1987. Television Culture. London: Methuen.
Gayo-Cal, Modesto, Mike Savage and Alan Warde. 2006. “A Cultural Map of the United
Kingdom, 2003.” Cultural Trends 15:213-37.
Gilula, Zvi and Shelby J. Haberman. 2001. “Analysis of Categorical Response Profiles by
Informative Summaries.” Sociological Methodology 31: 129-187.
Hirsch, Paul. 1972. “Processing Fads and Fashions: An Organizational Set Analysis of Cultural
Industry Systems.” American Journal of Sociology 77:639-659.
Katz-Gerro, Tally. 2002. “Highbrow cultural consumption and class distinction in Italy, West
Germany, Sweden, and the United States.” Social Forces 81:207–29.
------. 2006. “Comparative Evidence of Inequality in Cultural Preference: Gender, Class, and
Family Status.” Sociological Spectrum 26:63-83.
Katz-Gerro, Tally, Sharon Raz and Meir Yaish. 2007. “Class, Status, and the Intergenerational
Transmission of Musical Tastes in Israel.” Poetics 35:152-67.
Kraaykamp, Gebert, Koen van Eijck, Wout Ultee, and Kees van Rees. 2007. “Status and
Media Use in the Netherlands: Do Partners Affect Media Tastes?” Poetics 35:132-51.

33
Kuipers, Giselinde. 2006. “Television and Taste Hierarchy: The Case of Dutch Television
Comedy.” Media, Culture & Society 28:359-78.
Lambert, Paul, Manfred Max Bergman and Ken Prandy. 2005. “Leisurely Moments or
Lifetimes? Contexts and the Study of Leisure, Consumption, and Stratification.” Presented at the British Household Panel Survey-2005 Research Conference, Colchester, UK.
Levine, Lawrence W. 1984. “William Shakespeare and the American People: A Study in
Cultural Transformation.” American Historical Review 89: 34-66.
Lopes, Paul D. 1992. “Innovation and Diversity in the Popular Music Industry, 1969 – 1990.
American Sociological Review, 57:56-71.
López-Sintas, Jordi and Ercilia García-Álvarez. 2006. “Patterns of Audio-Visual Consumption:
The Reflection of Objective Divisions in Class Structure.” European Sociological Review
22:397-411.
López-Sintas, Jordi and Tally Katz-Gerro. 2005. “From exclusive to inclusive elitists and further: Twenty years of omnivorousness and cultural diversity in arts participation in the
USA.” Poetics 33:299-319.
Negus, Keith. 1999. Music Genres and Corporate Cultures. New York: Routledge.
Peterson, Richard A. 1972. “A process model of the folk, pop, and fine art phases of jazz.” Pp.
135-51 in American Music: From Storyville To Woodstock, edited by Charles Nanry. New
Brunswick, N.J.: Trans-Action Books and E.P. Dutton.
______. 1976. The Production of Culture. Beverly Hills, CA: Sage.

34
_____. 1979. “Revitalizing the Culture Concept.” Annual Review of Sociology 5:137-166.
_____. 1990. “Why 1955? Explaining the Advent of Rock Music. Popular Music 9:97-116.
_____. 1992. “Understanding audience segmentation: From elite and popular to omnivore and univore.” Poetics 21:243–58.
———. 1997. “The rise and fall of highbrow snobbery as a status marker.” Poetics 25:75–92.
_____. 2000. “The Ways Culture is Produced.” Poetics, 28:225-233.
———. 2005a. “Problems in comparative research: The example of omnivorousness.”
Poetics 33:257-282.
_____. 2005b. “Changing Arts Audiences: Capitalizing on Omnivorousness” Workshop
Paper: Cultural Policy Center at the University of Chicago.
_____ and Narasimhan Anand. 2004. “The Production of Culture Perspective.” Annual
Review of Sociology 30:311-334.
Peterson, Richard and David Berger. 1975. “Cycles in Symbol Production: The Case of
Popular Music.” American Sociological Review 40: 158 – 173.
_____ and Gabriel Rossman. 2007. “Changing Arts Audiences: Capitalizing on
Omnivorousness.” In Engaging Art: The Next Great Transformation of America’s Cultural Life, edited by William Ivey and Stephen J. Tepper. New York: Routledge. Forthcoming.
_____ and Roger M. Kern. 1996. “Changing highbrow taste: From snob to omnivore.”
American Sociological Review 61:900-907.

35
_____ and Albert Simkus. 1993. “How Musical Tastes Mark Occupational Status Groups.”
Pp. 152-186 in Cultivating Differences: Symbolic Boundaries and the Making of Inequality, edited by Michèle Lamont and Marcel Fournier. Chicago, IL: Chicago University Press.
Sonnett, John. 2004. “Musical Boundaries: Intersections of Form and Content.” Poetics 32:
247-264.
Storey, John. 2003. “The Social Life of Opera.” European Journal of Cultural Studies 6:5-35.
Torche, Florencia. 2007. “Social Status and Cultural Consumption: The Case of Reading in
Chile.” Poetics 35:70-92. van Eijck, Koen. 1999. “Socialization, Education, and Lifestyle: How Social Mobility Increases the Cultural Heterogeneity of Status Groups.” Poetics 26:309-328.
———. 2001. “Social differentiation in musical taste patterns.” Social Forces 79: 1163-1185. van Eijck, Koen and Bertine Bargeman. 2004. “The changing impact of social background on lifestyle: “culturalization” instead of individualization?” Poetics 32: 447-469 van Eijck, Koen and Wim Knulst. 2005. “No More Need for Snobbism: Highbrow Cultural
Participation in a Taste Democracy.” European Sociological Review 21:513-28. van Eijck, Koen and Kees van Rees. 2000. “Media Orientation and Media Use: Television
Viewing Behavior of Specific Reader Types from 1975 to 1995.” Communication Research
27:574-616.

36 van Rees, Kees, Vermunt, Jeroen K. and Marc Verboord. 1999. “Cultural Classifications
Under Discussion: Latent Class Analysis of Highbrow and Lowbrow Reading.” Poetics 26:
349-365.
Vissol, Terry. 2006. Is There a Case for an EU Television Station? Luxembourg: Office for the
Official Publications of the European Communities.
Warde, Alan, David Wright and Modesto Gayo-Cal. 2007. “Understanding Cultural
Omnivorousness: Or, the Myth of the Cultural Omnivore.” Cultural Sociology 1:143-64.
Williams, Raymond. 1973. “Base and Superstructure in Marxist Cultural Theory.” New Left
Review 82 (October).
Wright, David. 2006. “Cultural Capital and the Literary Field.” Cultural Trends15:123-39.

37
.75
.5
r = -.74
portugal
.25
ireland
0
-.25
-.5
belgium sweden greece netherlands luxembourg denmark austria united kingdom germany finland spain italy france
-.75
0 2 4 6
Average Market-Penetration Rank Score
8
Figure 1. Scatterplot of the association between the highbrow/non-highbrow difference in television consumption omnivorousness and the relative degree of market-orientation of television production for 15 European countries, 2001 Eurobarometer.

38
Table 1. Sociodemographic characteristics of respondents by patterns of highbrow consumption.
Pattern Classical Opera Jazz College Occupation a Urban Museum Age %
1
2
3
4
5
6
7
8
Total
+
+
+
-
-
-
-
+
-
-
+
+
-
-
+
+ a Managerial or Professional.
-
+
-
+
-
+
-
+
17.2%
30.0%
16.4%
30.6%
33.9%
24.8%
17.0%
29.5%
13.6%
20.8%
24.1%
47.5% 34.8%
35.8% 21.4%
50.1% 29.4%
20.8%
28.1% 22.4%
34.6% 41.0%
22.7% 35.5%
33.3% 44.4%
27.8% 44.3%
40.9
35.8
57.6
52.7
49.3 14.67
35.0% 59.1% 43.0 6.51
31.6% 50.4% 58.5 5.33
37.5% 66.5% 49.7 3.65
29.4% 33.0% 43.6
58.99
8.25
2.16
0.43

39
Table 2. Differences between highbrows and non-highbrows in omnivore music consumption, 2001, Eurobarometer.
Omnivore Score (Music)
Belgium
Denmark
Germany
Greece
Italy
Spain
France
Ireland
Luxembourg
Netherlands
Portugal
United Kingdom
Finland
Sweden
Austria
Highbrow Other Difference
2.47
2.25
1.73
3.09
1.78
2.33
2.33
1.84
2.39
1.80
3.73
2.33
2.73
2.37
1.34
1.82
2.23
1.98
1.33
1.44
1.58
1.88
1.53
2.13
2.03
1.69
1.83
2.30
2.24
1.82
0.65
0.02
-0.25
1.76
0.34
0.75
0.45
0.31
0.25
-0.23
2.04
0.50
0.44
0.14
-0.48
F p-value
23.04
0.03
7.00
0.00
0.86
0.01
124.99 0.00
5.79 0.02
23.85
12.13
0.00
0.00
5.81
2.80
0.02
0.09
3.95
117.31
24.36
10.56
1.57
24.47
0.05
0.00
0.00
0.00
0.21
0.00

40
Table 3. Differences between highbrows and non-highbrows in omnivore television consumption, 2001, Eurobarometer.
Omnivore Score (TV)
Belgium
Denmark
Germany
Greece
Italy
Spain
France
Ireland
Luxembourg
Netherlands
Portugal
United Kingdom
Finland
Sweden
Austria
Highbrow
3.39
2.81
3.02
2.80
2.88
2.39
2.43
3.23
3.23
3.47
3.50
3.26
3.42
3.43
2.61
Other Difference
3.47
3.08
3.18
2.71
3.01
2.88
3.07
3.04
3.42
3.46
3.01
3.57
3.86
3.50
2.85
-0.08
-0.27
-0.16
0.09
-0.14
-0.49
-0.64
0.19
-0.19
0.01
0.50
-0.30
-0.44
-0.06
-0.24
F p-value
0.19
6.44
1.77
0.22
0.91
0.67
0.01
0.18
0.64
0.34
6.59 0.01
18.06 0.00
1.12
1.36
0.29
0.24
0.01
4.63
5.26
7.13
0.92
0.03
0.02
0.01
0.31
2.73
0.58
0.10

41
Table 4. Random-effects regression models of the effect of individual sociodemographic characteristics, highbrow status and national television production regime on the number of different types of television programs watched, 2001 Eurobarometer.
Highbrow
Market Orientation of Television Field
Highbrow X Market Orientation
Age
Age Squared
Education (College=1)
Gender (Female=1)
Marital Status (Married=1)
Occupation (Keeping House=1)
Employment Status (Retired=1)
Occupation (Manual=1)
Constant
χ 2
σ(μ j
)
N
*p<0.05
Model 1
-0.1952**
(-5.00)
3.2058**
(40.14)
24.97**
0.30**
16200
Model 2
**p<0.01 (two-tailed test; t-statistics in parentheses)
-0.1979**
(-5.07)
0.0113
(0.25)
-0.0951**
(-3.94)
3.2057**
(37.09)
40.56**
0.33**
16200
Model 3
-0.0139
(-0.36)
0.0176
(0.40)
-0.0887**
(-3.76)
-0.0556**
(-13.03)
0.0004**
(8.29)
-0.2445**
(-7.48)
0.3023**
(11.03)
0.0663*
(2.25)
0.2677**
(5.59)
0.3281**
(6.68)
0.1445**
(4.23)
4.5045**
(37.93)
877.74**
0.32**
16200