You are on page 1of 12

Marketing Letters 16:2, 7586, 2005

c 2005 Springer Science + Business Media, Inc. Manufactured in the Netherlands.




The Role of Ordinary Evaluations in the Market


for Popular Culture: Do Consumers Have Good
Taste?
MORRIS B. HOLBROOK
Graduate School of Business, Columbia University, New York, NY 10027, USA

mbh3@columbia.edu

Abstract
If we define good taste as that prescribed by professional experts in a particular cultural field and ask whether
ordinary consumers (non-experts or members of the mass audience) have good taste, the evidence from previous
studies suggests that the relationship between expert judgments and popular appeal to ordinary consumers is
significantly but only weakly positive and is therefore consistent with a phenomenon of little taste. Possible
explanations stem from the consideration of a variable that might mediate and thereby weaken the relationship
between expert judgments and popular appealnamely, ordinary evaluations, in which non-expert consumers
assess the excellence (rather than the enjoyability) of a cultural offering. An earlier experimental study of musical
performances showed that ordinary evaluations did intervene between expert judgments and popular appeal to
college students so that, in this sense, ordinary consumers did display aspects of good taste. New data on over 200
motion pictures corroborate this finding in another cultural context, with actual audience members, and through
the use of real-world as opposed to experimental observations.
Keywords: expert judgments, ordinary evaluations, popular appeal, professional critics, mass audiences, motion
pictures, consumer tastes

1.

Introduction

For over two thousand years, a debate has raged on the theme of whether consumers in
general or audiences in particular display good taste in their appreciation of cultural
offerings (Bayley, 1991; Brantlinger, 1983; Ross, 1989; Strinati, 1995; Twitchell, 1992;
Zolberg, 1990). This debate has fostered unresolved tensions between ideological antagonists ranging from those who wage elitist defenses of artistic integrity to those, at the
other extreme, who voice their liberal egalitarian tendencies in the form of strident political correctness. Fortunately, such thankless polemicswhich generally exist in a vacuum
of empirical evidencefall outside the scope of the present paper. Specifically, this paper
pursues data-based answers to what have too often been treated as ideology-based questions.
1.1.

The Post-Bourdieusian Perspective

As suggested by the French sociologist Pierre Bourdieu (1983, 1984, 1986, 1993), a phenomenon occurs within a given cultural field (music, films, . . . ) where, by virtue of cultural

76

HOLBROOK

capital acquired through long training and demonstrated by acknowledged expertise, certain
individuals (professionals, scholars, . . . ) are recognized and legitimated as arbiters of good
taste in that particular cultural field (musicologists, cinema critics, . . . ). This perspective
raises the question of whether the tastes of ordinary or non-expert consumers tend to reflect
those of expert professionals who, after prolonged study, know what is judged to be good
according to the criteria sanctioned by the relevant cultural field. Bourdieu himself has
tended to assume that the answer is negative (1983, p. 330; 1984, p. 271; 1986, p. 152;
1993, p. 164), but this issue remains in need of empirical testing.
1.1.1. Resolving the debate? The Post-Bourdieusian Perspective invites empirical investigation to determine whether the relationship between judgments by professional experts and those by ordinary consumers in a particular cultural field is, in fact, (1) negative
(bad taste); (2) positive (good taste); or (3) something in between (no taste or little
taste).
1.1.2. The evidence On balance, the rather sparse empirical evidence that has addressed
this question appears to favor a something-in-between conclusion (no taste or little
taste). Specifically, just a few scattered studies in various cultural fields have collected
actual data on the relationship between expert judgments by trained professionals and popular appeal to ordinary audience members. These studies have sometimes found small
negative correlations (e.g., Hirschman and Pieros, 1985); have sometimes reported nonmonotonic (U-shaped) relationships (Wallace et al., 1993); and have sometimes shown no
relationships (Ravid, 1999); but have most commonly demonstrated statistically significant but only weakly positive associationsthat is, correlations (r ) or standardized beta
coefficients () on the general order of .06 < r -or- < .36between popular appeal
and (1) journalistic reviewers (Eliashberg and Shugan, 1997; Basuroy et al., 2003); (2)
trained experts (Schindler et al., 1989; Holbrook et al., 2003); or (3) professional critics (Holbrook, 1999, 2003). Apparently, if we define good taste as that legitimated
by professional experts in a particular cultural field, then ordinary consumers do not
tend to display good taste thus defined. Rather, they display little taste in the sense
of showing only a significant-but-weak positive relationship of popular appeal to expert
judgments.
1.2.

Two Competing Explanations

All this raises the obvious question of why ordinary or non-expert consumers tend to
display less-than-good or, indeed, little taste. In this connection, we might consider
two competing explanations that involve assumptions about a previously-neglected but
potentially-important intervening variablenamely, Ordinary Evaluation via assessments
of excellence by non-expert consumers (cf. Charters and Pettigrew, 2005). Though ignored by all but one previous study, Ordinary Evaluation (assessments of excellence by
non-expert consumers) differs from both Expert Judgment (assessments of excellence by

THE ROLE OF ORDINARY EVALUATIONS IN THE MARKET FOR POPULAR CULTURE

Expert
Judgment

Ordinary
Evaluation

near 0

77

Popular
Appeal

Figure 1. Dignity-of-the-Common-Person Hypothesis (DCPH).

professional critics) and Popular Appeal (expressions of liking by non-expert consumers).


Hypothetically, Ordinary Evaluation mediates and thereby weakens the relationship between Expert Judgment and Popular Appeal, helping to account for the aforementioned
phenomenon of little taste. Let us therefore look closely at each of these two competing
explanations.
1.2.1. Competing Explanation #1: The Dignity-of-the-Common-Person Hypothesis
(DCPH) First, consistent with the egalitarian views of many liberal thinkers, Competing
Explanation #1the Dignity-of-the-Common-Person Hypothesis (DCPH)assumes that
people share the norms for what is considered good by those with expertise in a particular
cultural field and thereby display aspects of good taste, but nonetheless may tend to like
something regarded by experts as bad just because it is more enjoyable, pleasant, exciting, fun, accessible, or easy to understand. Diagrammatically, as shown in Figure 1, DCPH
explains the oft-observed significant-but-weak positive correlation between Expert Judgment and Popular Appeal (the little taste phenomenon) by a positive association between
Expert Judgment and Ordinary Evaluation (aspects of good taste) that is dissipated by
a near-zero relationship between Ordinary Evaluation and Popular Appeal (aspects of no
taste).
1.2.2. Competing Explanation #2: The What-a-Wonderful-World Hypothesis (WWWH)
Second and by contrast, Competing Explanation #2the What-a-Wonderful-World Hypothesis (WWWH)recalls the old song of the same title by Sam Cooke: Dont know
much about history . . . biology. . . geography, but do know whom I love. WWWH assumes
that people have no idea what is considered good or bad by the expert-legitimated
norms for excellence in a particular cultural fieldin other words, dont know much
about (say) music . . . cinema . . . whateverbut do know what they like and sometimes or
even often happen to like what the professional critics regard as inferior. Diagrammatically,
as shown in Figure 2, WWWH explains the oft-observed significant-but-weak positive correlation between Expert Judgment and Popular Appeal (the little taste phenomenon) via
a near-zero association between Expert Judgment and Ordinary Evaluation (aspects of no

Expert
Judgment

near 0

Ordinary
Evaluation

Figure 2. What-a-Wonderful-World Hypothesis (WWWH).

Popular
Appeal

78

HOLBROOK

Expert
Judgment

.548

Ordinary
Evaluation

.592

Popular
Appeal

.040 (NS)
Figure 3.

Mediating role of Ordinary Evaluation in the case of musical performances (Holbrook et al., 2003).

taste) that cannot be overcome by any positive relationship that might or might not exist
between Ordinary Evaluation and Popular Appeal (potential aspects of good taste).
1.3.

Previous Findings

Earlier studies have typically not measured Ordinary Evaluation and, therefore, have not
presented data suitable for adjudicating between these two competing explanations, DCPH
versus WWWH. However, one recent study by Holbrook et al. (2003) did investigate this
issue experimentally in the case of music. Specifically, 200 recorded performances of My
Funny Valentine were assessed by independent samples for (1) Expert Judgment (EJ)
via ratings of aesthetic excellence by highly trained professional musicians; (2) Ordinary
Evaluation (OE) via ratings of aesthetic excellence by ordinary non-expert consumers (college students with no special musical education); and (3) Popular Appeal (PA) via ratings
of enjoyability by a separate randomly-assigned group of ordinary non-expert consumers
(untrained college students). The overall relationship between EJ and PA of rPA,EJ = .365
compares closely with the significant-but-weak positive associations (little taste) reported
in previous studies, none of which included any measure of OE. Moreover, as indicated by
Figure 3, considering OE as a potential mediator showed partial support for DCPH (aspects
of good taste) in that OE did reflect EJ (rOE,EJ = .548). However, consistent with WWWH
(potential aspects of good taste), PA reflected OE (PA,OE.EJ = .592). Hence, the overall
correlation between EJ and PA remained significantly but weakly positive (the phenomenon
of little taste): rPA,EJ = (.548 .592) + .040 = .365.
1.4.

Preview

The purpose of the present inquiry is to extend previous investigations by further exploring the potential mediating role of ordinary evaluations as a possible explanation for the
often-documented significant-but-weak relationship between expert judgments and popular
appeal in the appreciation of cultural offerings. Herein further examining the intervening
role of ordinary evaluations, formerly investigated only in a laboratory experiment with
college students evaluating performances of one songthe present study enlarges its scope
to cover a broader level of application to another major type of cultural offering by addressing the case of motion pictures in the context of real-world assessments by members of the
mass audience.

THE ROLE OF ORDINARY EVALUATIONS IN THE MARKET FOR POPULAR CULTURE

2.
2.1.

79

Method
Sample

For comparability with the sample size of 200 performances in the My Funny Valentine
study, the present investigation used the 229 feature-length motion pictures released during
the year 2000as reported by www.worldwideboxoffice.com, which gives figures for the
box-office revenues of each film. Due to missing data on other variables, the final sample
size was reduced by 10 to N = 219 films.
2.2.

Measures

On these 219 films, as now described, the present study collected measures of three main
typesintended to extend the findings of the music-related My Funny Valentine study,
conducted in an experimental laboratory setting and based on the responses of college
students, to the case of motion pictures in a real-world setting based on responses by
members of the movie audience.
2.2.1. Expert Judgment The measure of Expert Judgment came from six books providing professional assessments of excellence by various established critical authorities
who claim to be concerned with the degree of a films long-term merit as an enduring
contribution to popular culture or its lack thereof: (1) Maltin (2002)seven-position ratings in half-step increments from one star = a bomb to four stars = the very best; (2)
Martin and Porter (2002)nine-position ratings in half-step increments from one star = a
turkey to five stars = excellent; (3) TV Guide (2002, supplemented where necessary by
www.tvguide.com/movies/database)nine-position ratings in half-step increments from
one star = without merit to five stars = a masterpiece; (4) Craddock (2002)nineposition ratings in half-step increments from zero bones = Woof! to four bones = the
highest rating; (5) Walker (ed. 2002)nine-position ratings in half-step increments from
zero stars = a totally routine production or worse to four stars = outstanding in many
ways, a milestone in cinema history, remarkable for acting, direction, writing, photography or some other aspect of technique; and (6) Ebert (2002)seven-position ratings in
half-step increments from one star = poor to four stars = a great film. These six scores
based on expert judgments of excellence by six professional critics were each standardized
across films. The resulting z-scores were then summed to form a six-item index of Expert
Judgment. The reliability of this six-item index was quite satisfactory: Mean inter-item
correlation = .474; mean corrected item-total correlation = .623; coefficient alpha = .844
(not improved by eliminating any of the six items). As a partial check on its validity, this
six-item index of Expert Judgment was correlated with the average ratings reported by
www.rottentomatoes.com, which tabulates the gist of movie reviews from a large pool of
journalistic reviewers. Average scores from reviewers at Rotten Tomatoes (RT) showed a
strong association with those for Expert Judgment (EJ), suggesting a satisfactory degree of
convergent validity for the measure used here: rEJ,RT = .893 (t217 = 29.304, p < .001).

80

HOLBROOK

2.2.2. Ordinary Evaluation The measure of Ordinary Evaluation came from the Internet Movie Database at www.imdb.com. On this IMDb Website, registered members
from the general populace rate various films on a ten-point scale from 1 = awful to 10
= excellent. These responses appear as a weighted-average rating for excellence
weighted in the sense of filtering to eliminate or reduce any potential effects of ballot
stuffing.
2.2.3. Popular Appeal The measure of Popular Appeal also came from www.imdb.com.
Here, we need to make an assumption concerning what is, in effect, an indirect measure of
popularitynamely, a measure based on the number of people who provide IMDb ratings
of a given film. This statistic can be viewed in various ways, most of them consistent with an
interpretation as the degree to which more/fewer ordinary consumers or audience members
like, enjoy, or praise a particular film by recommending it to others. For example, according
to a seemingly reasonable popularity hypothesis, the number providing IMDb ratings of
a film would indicate the level of appreciation or enthusiasm it produces; the degree of
favorable attention it attracts; word of mouse as a parallel to word of mouth; popular
buzz; and/or the general desire to recommend the film. According to this intuitively
plausible popularity hypothesis, the number of IMDb votes should increase with the
degree to which people enjoy a film (where people like and recommend films they consider
good). Nonetheless, an alternative extremity hypothesis might suggest that more people
would vote for a film on the IMDb Website the more strongly they like it but also the more
strongly they dislike it. According to this competing extremity hypothesis, the absolute
distance of ratings from their mean should determine the number of IMDb votes (where
people like and recommend films they consider good but also dislike and warn against
films they consider bad).
The comparative validity of these two hypotheses was assessed in three ways. First,
the number of IMDb votes (NV) across films was correlated with their overall ratings
or Ordinary Evaluation (OE) to estimate rNV,OE (the popularity hypothesis) and with the
absolute distances of those ratings from their mean (AD) to estimate rNV,AD (the extremity
hypothesis). Consistent with the popularity hypothesis, the former correlation greatly
exceeded the latter in strength: rNV,OE = .363 (t217 = 5.737, p < .001) versus rNV,AD =
.155 (t217 = 2.318, p = .02). From a second perspective, for films at or above the mean
in OE and consistent with the popularity hypothesis, rNV,OE = .486 (t118 = 6.037, p <
.001); meanwhile, for those below the mean in OE and still supportive of the popularity
hypothesis versus the extremity hypothesis, this correlation remains positive at rNV,OE =
.231 (t97 = 2.338, p = .02). From a third viewpoint, the extremity hypothesis was further
assessed by checking for a possible non-monotonic relationship between NV and OE via
2
the inclusion of a linear term (z OE , the standardized value for OE), a quadratic term (z OE
),
3
2
3
and a cubic term (z OE ) in a regression of NV on z OE , z OE , and z OE . Here, the overall fit
was fairly strong (R = .557, F3,215 = 32.211, p < .001), with significant contributions
2
from z OE ( = .211, t215 = 2.490, p = .01), z OE
( = .580, t215 = 7.449, p < .001), and
3
z OE ( = .511, t215 = 4.851, p < .001) such that the shape of the relationship between OE
and NV was comparatively flat at low levels of OE and increased upward at high levels of
OE. All this strongly supports the popularity hypothesis at the expense of the extremity

THE ROLE OF ORDINARY EVALUATIONS IN THE MARKET FOR POPULAR CULTURE

81

hypothesis and thereby suggests the validity of the number of IMDb votes as a measure of
popular appeal.
In the present sample of films, as a representation of popular appeal, the numbers of IMDb
votes (NV) ranged from 71 to 60,492with all but 2/3/8/14/23 > 100/200/300/400/500,
respectively (Mean = 5,821, SD = 8,006.79, SE = 541.05, Median = 2,710). As can be
seen by comparing the mean and median, scores for NV were skewed toward the high
end of the distribution. For this reason and because the relationship between NV and OE
showed signs of curvilinearity as well as herteroscedasticity, a logarithmic transformation
of NV was introduced to bring its mean and median into closer alignment, to flatten out its
relationship with OE, and to avoid problems of heteroscedasticity (Mean = 3.4480, SD =
.5553, SE = .0375, Median = 3.4330).
As a partial check on the validity of Popular Appeal (PA) defined as log10 NV, PA was
correlated with logarithmically-transformed measures of domestic box office (DBO) and international box office (IBO) taken from www.worldwideboxoffice.com (where rDBO,IBO =
.572, t217 = 10.285, p < .001). The resulting correlations supported the convergent validity
of the logarithmically-transformed vote-based measure of Popular Appeal by indicating, as
would be expected, that PA shares variance with logarithmically-transformed box-office revenues: rPA,DBO = .677 (t217 = 13.542, p < .001) and rPA,IBO = .607 (t217 = 11.251, p <
.001).

2.3.

Rationale

Notice thatas in other film-oriented research reviewed earlier (e.g., Basuroy et al., 2003;
Eliashberg and Shugan, 1997; Holbrook, 1999; Ravid, 1999; Wallace et al., 1993)the
method followed here pursues an approach comparable to that sometimes employed in
research on advertising (Holbrook and Batra, 1987; Olney et al., 1991). Specifically, in the
manner recommended by Jackson (1992), the present measures and tests are conducted
across objects (i.e., motion pictures) rather than across people (i.e., professional experts or
ordinary non-expert consumers). This approach carries both advantages and disadvantages.
As an advantage, measures are collected from independent samples (e.g., professional
experts versus ordinary consumers), thereby reducing the dangers of methods artifacts such
as consistency biases that might occur when a respondent answers multiple questions in
a manner governed by logical coherence or response-style biases due to yea-/nay-saying
tendencies that cut across multiple answers. Furtherunlike typical experimental designs,
which often use highly restricted forms of comparison between (say) Messages A and B or
Objects 1 and 2the sample of cultural offerings includes over 200 films, thereby helping
to insure that results do not hinge on peculiarities of particular items sampled.
As a disadvantage, the present approach does not permit behavioral interpretations along
the lines of concluding that someone who reads a favorable review tends to form a positive
evaluation and therefore leans toward enjoying the film. Rather, we ask whether a film of the
type that prompts favorable expert judgments of excellence tends also to elicit evaluations
as excellent by ordinary consumers and/or whether a film of the sort judged excellent by
consumers tends to win a high level of popular appeal.

82

HOLBROOK

In other words, the present measures of expert judgment, ordinary evaluation, and popular
appeal are not intended to represent a temporal sequence of effects that unfolds over time
at the individual level of analysis. Rather, they provide cross-sectional data, across films
released during a given year. Specificallybecause these measures are taken at a moment
in time, in this case roughly three years after the release of a film, with ordinary evaluation
and popular appeal measured cumulatively at a time approximately contemporaneous with
the measure of expert judgmentthey coincide at a moment that occurs after the film has
finished its theatrical run and after interested members of the audience have had a chance
to watch it in the theaters and/or on pay-per-view television, cable TV, videotape, or DVD.
Hence, the present measures reflect assessments of a film viewed as a cultural offering for
the ages and not the market dynamics of its promotional campaign (advertising, trailers,
guest appearances on talk shows, etc.) or distribution strategy (number of screens, opening
box office, timing of video release, etc.). These marketing-related aspects of a films release
are no doubt critical to its box-office success and are therefore of considerable potential
interest to marketing researchers. However, our present focus deals with aspects of a film
whereby its evaluations by professional critics and ordinary consumers become part of its
long-term assessment as a cultural object whose positive or negative appraisal persists over
time as an enduring contribution to popular culture.
3.

Results

The present results show an overall pattern quite consistent with the results from several
studies reviewed earliernamely, a significant-but-weak positive correlation between Expert Judgment (EJ) and Popular Appeal (PA): rPA,EJ = .225 (t217 = 3.396, p < .001).
Further, the present findings for a real-world sample of films mirror those obtained
previously for the case of musical excerpts evaluated by college students in a laboratory
experiment. Specifically, as shown by the path analysis diagrammed in Figure 4, the overall
relationship between Expert Judgment (EJ) and Popular Appeal (PA) can be explained
via the mediating role of Ordinary Evaluation (OE) as an intervening variable (Baron and
Kenny, 1986; Pedhazur, 1982):
rOE,EJ = .841 (t217 = 22.921, p < .001);
rPA,OE = .297 (t217 = 4.582, p < .001);
PA,OE.EJ = .370 (t216 = 3.079, p = .002);
PA,EJ.OE = .086 (t216 = .719, NS);
rPA,EJ = (.841 .370) .086 = .225 (t217 = 3.396, p < .001).
Expert
Judgment

.841

Ordinary
Evaluation

.370

Popular
Appeal

-.086 (NS)

Figure 4.

Path model including Ordinary Evaluation in the case of motion pictures from the year 2000.

THE ROLE OF ORDINARY EVALUATIONS IN THE MARKET FOR POPULAR CULTURE

4.
4.1.

83

Discussion
Summary of Key Findings

These results corroborate and extend earlier findings from the My Funny Valentine study
obtained in the context of a laboratory experiment in which college students evaluated
musical recordings, where ordinary evaluations played an intervening role in mediating the
relationship between expert judgments and popular appeal. The present study extends these
findings to the case of motion pictures evaluated by real-world consumers and supports the
validity of the Dignity-of-the-Common-Person Hypothesis (DCPH).
First, supporting DCPH and contradicting WWWH, ordinary evaluations do reflect expert
judgments (r = .841, t217 = 22.921, p < .001, in Figure 4). In other words, consistent
with DCPH, evaluations of films by ordinary consumers do reflect what expert critics regard
as more or less excellent (aspects of good taste).
Second, as in the My Funny Valentine study, the results also indicate a weak-butpositive connection between expert judgments and popular appeal in the case of motion
pictures (the little taste phenomenon). Specifically, popular appeal does reflect expert
judgments to a significant-but-weak degree (r = .225, t217 = 3.396, p < .001).
Third, this relationship [(.841 .370) .086 = .225, in Figure 4] is mediated by the
intervening role of ordinary evaluations (PA,OE.EJ = .370, t216 = 3.079, p = .002), with
just a small and statistically insignificant direct contribution of expert judgments to popular
appeal when controlling for ordinary evaluations (PA,EJ.OE = .086, t216 = .719, NS).
Thus, consistent with DCMH, the mediating role of OE explains how the fairly strong
rOE,EJ = .841 (aspects of good taste) is diluted by the much weaker PA,OE.EJ = .370
(aspects of little taste) to explain the overall weak rPA,EJ = .225 (the little taste phenomenon).

4.2.

Limitations and Directions for Future Research

Ironically, the major limitation of the present study stems from its major advantage. Specifically, across a large sample of films, this study has obtained independent measures of expert
judgments and ordinary evaluations plus popular appeal, thereby guarding against artifacts
due to shared-methods biases. Drawing on this advantage, a path analysis has allowed us
to identify a mediating relationship wherein Expert Judgment Ordinary Evaluation
Popular Appeal. However, we must exercise some caution in interpreting these findings.
Given the present approach, we are in a positionat the aggregate level of analysis and
with films as the units of observationto claim (say) that films esteemed by professional
critics tend to inspire positive evaluations by ordinary consumers and that films with positive
ordinary evaluations tend to generate high levels of popular appeal in the form of liking,
attention, and word of mouth or click of mouse. But we must interpret these findings with
caution because they do not permit tempting behavioral inferencesat the individual level
of analysis and with people as the units of observationto the effect (say) that ordinary
consumers base their evaluations of excellence on the expert judgments of professional

84

HOLBROOK

critics or that ordinary evaluations exert a causal impact on popular appeal in the form
of liking, attention, and recommendations to others. Rather, behaviorally, it could be that
expert critics and ordinary consumers both respond to the same movie-related features
(cinematography, acting, directing, etc.) or that ordinary evaluations and popular appeal
both reflect marketing-related aspects of a films promotional strategy (advertising, trailers,
wide or narrow theatrical distribution, appearances on television, availability on video,
special deals on www.amazon.com, etc.).
Put differently, in the manner recommended by Jackson (1992) and illustrated elsewhere in the context of advertising (Holbrook and Batra, 1987; Olney et al., 1991), the
present measures and tests are conducted across films rather than across people. Thus, the
study shows not that consumers who read a positive/negative critical appraisal tend to form
favorable/unfavorable evaluations of that film (across people, as would be typical in an
experimental study), but rather that the sorts of movies that win positive expert judgments
tend to prompt favorable consumer evaluations (across films, via a correlational analysis).
Similarly, the study shows not that consumers who regard a film as excellent tend to enjoy
it (across people, as in an experiment), but rather that the sorts of movies that ordinary consumers regard as excellent tend to promote popular appeal in the form of liking, attention,
or word of mouse (again, across films, correlationally).
In sum, this aggregate-level across-films correlational approachas compared with an
individual-level across-people experimental approach-has certain disadvantages but also
compensating advantages (Jackson, 1992). As a disadvantage, it limits our interpretations
behaviorally in that we cannot make inferences concerning the influence process at the
individual level (e.g., from expert critics to ordinary consumers) or the causal determinants
of popular appeal to specific consumers (e.g., from evaluations of excellence to enjoyment
or buzz). As a compensating advantage, however, the present measures and tests permit
assessing a communication chain across films wherein the sort of movie that tends to be
critically praised by experts also tends to be evaluated favorably by ordinary consumers
(aspects of good taste) and, to a lesser extent, to win popular appeal (the little taste
phenomenon).
All this suggests a desirable direction for future researchnamely, experimental or perhaps ethnographic studies to investigate how individuals use expert judgments by professional critics to form evaluations concerning the merits of motion pictures and how those
evaluations of merit by ordinary consumers shape preferences and tendencies to recommend
films to others. In other words, future studies should undertake individual-level investigations (across people) to supplement the present aggregate-level findings (across films).

4.3.

Conclusions

Subject to these and other potential limitations that indicate the need for further research,
it appears that the present findings permit some fairly firm conclusions concerning the
questions with which we began.
Specifically, in the case of motion pictures evaluated in a real-world setting, it appears thatconsistent with the Dignity-of-the-Common-Person Hypothesis (DCPH) and

THE ROLE OF ORDINARY EVALUATIONS IN THE MARKET FOR POPULAR CULTURE

85

reflecting aspects of good tastefilms judged as high in worth by professional critics


tend to receive positive evaluations from ordinary consumers. However, consistent with
the significant-but-weak or little taste phenomenon, the relationship between ordinary
evaluations and popular appeal is only moderately positive so that the overall relationship
between expert judgments and popular appeal remains quite modest.
This train of thoughtby virtue of including an explicit recognition of the mediating role
played by ordinary evaluationsgoes some distance toward resolving or at least refining
the debate that has long raged among various commentators on popular culture. In the end,
it appears that both sides are partially right. In the case of motion pictures, at the aggregate
level of analysis, consumers do know whats considered good by experts. Further, to a
lesser degree, popularity does tend to reflect these consumer evaluations of excellence. But
the combined multiplicative relationship between expert judgments and popular appeal,
though significant, remains weak. In short, in their ordinary evaluations, consumers display
aspects of good taste; but, in registering popular appeal, they show little taste overall.

Acknowledgment
The author gratefully acknowledges the support of the Columbia Business Schools Faculty
Research Fund.

References
Baron, Reuben M., and David A. Kenny. (1986). The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations, Journal of Personality and Social
Psychology 51(December), 11731182.
Basuroy, Suman, Subimal Chatterjee, and S. Abraham Ravid. (2003). How Critical Are Critical Reviews? The
Box Office Effects of Film Critics, Star Power, and Budgets, Journal of Marketing 67(October), 103117.
Bayley, Stephen. (1991). Taste: The Secret Meaning of Things. New York: Pantheon.
Bourdieu, Pierre. (1983). The Field of Cultural Production, Or: The Economic World Reversed, Poetics 12,
311356.
Bourdieu, Pierre. (1984). Distinction: A Social Critique of the Judgement of Taste. Cambridge: Harvard U. Press.
Bourdieu, Pierre. (1986). The Production of Belief: Contribution to an Economy of Symbolic Goods. In Richard
Collins et al. (eds.), Media, Culture and Society: A Critical Reader, London: Sage Publications, pp. 131163.
Bourdieu, Pierre. (1993). The Field of Cultural Production: Essays on Art and Literature, (ed.) Randal Johnson.
New York: Columbia U. Press.
Brantlinger, Patrick. (1983). Bread & Circuses: Theories of Mass Culture As Social Decay. Ithaca: Cornell U.
Press.
Charters, Stephen, and Simone Pettigrew. (2005). I Like It, But How Do I know if Its Any Good?:
Quality and Preference in Wine Consumption, Journal of Research for Consumers Issue 5, on-line
@www.jrconsumers.com.
Craddock, Jim. (2002). VideoHounds Golden Movie Retriever. Farmington, MI: Gale.
Ebert, Roger. (2002). Roger Eberts Movie Yearbook 2003. Kansas City, MO: Andrews McMeel Publishing.
Eliashberg, Jehoshua, and Steven M. Shugan. (1997). Film Critics: Influencers or Predictors? Journal of Marketing 61(April), 6878.
Hirschman, Elizabeth C., and Andrew Pieros, Jr. (1985). Relationships Among Indicators of Success in Broadway
Plays and Motion Pictures, Journal of Cultural Economics 9(June), 3563.

86

HOLBROOK

Holbrook, Morris B. (1999). Popular Appeal versus Expert Judgments of Motion Pictures, Journal of Consumer
Research 26(September), 144155.
Holbrook, Morris B. (2003). Perish Or Publish: On the Cultivation of Bad Taste and Surviving Your Fifteen
Minutes of Fame, Working Paper, Graduate School of Business, Columbia U.
Holbrook, Morris B., and Rajeev Batra. (1987). Assessing the Role of Emotions as Mediators of Consumer
Responses to Advertising, Journal of Consumer Research 14(December), 404420.
Holbrook, Morris B., Kathleen T. Lacher, and Michael S. LaTour. (2003). Popular Appeal Versus the Evaluations
of Ordinary Consumers and Critical Experts: An Illustration Based on Musical Recordings of My Funny
Valentine, Working Paper, Graduate School of Business, Columbia U.
Jackson, Sally. (1992). Message Effects Research. New York: Guilford Press.
Maltin, Leonard. (2002). Leonard Maltins Movie & Video Guide: 2003 Edition. New York: Plume.
Martin, Mick, and Marsha Porter. (2002). Video & DVD Guide 2003. New York: Ballantine Books.
Olney, Thomas J., Morris B. Holbrook, and Rajeev Batra. (1991). Consumer Responses to Advertising: The
Effects of Ad Content, Emotions, and Attitude toward the Ad on Viewing Time, Journal of Consumer Research
17(March), 440453.
Pedhazur, Elazar J. (1982). Multiple Regression in Behavioral Research: Explanation and Prediction, Second
Edition. New York: Holt, Rinehart and Winston.
Ravid, S. Abraham. (1999). Information, Blockbusters, and Stars: A Study of the Film Industry, Journal of
Business 72(4), 463492.
Ross, Andrew. (1989). No Respect: Intellectuals & Popular Culture. New York: Routledge.
Schindler, Robert M., Morris B. Holbrook, and Eric A. Greenleaf. (1989). Using Connoisseurs to Predict Mass
Tastes, Marketing Letters 1(1), 4754.
Strinati, Dominic. (1995). An Introduction to Theories of Popular Culture. London: Routledge.
TV Guide. (2002). TV Guide Film & Video Companion: Compiled By the Editors of TV Guide Onlines Cinebooks
Database. New York: Friedman/Fairfax.
Twitchell, James B. (1992). Carnival Culture: The Trashing of Taste in America. New York: Columbia U. Press.
Walker, John (ed.). (2002). Halliwells Film & Video Guide 2003, 18th Edition. New York: HarperResource.
Wallace, W. Timothy, Alan Seigerman, and Morris B. Holbrook. (1993). The Role of Actors and Actresses in the
Success of Films: How Much Is a Movie Star Worth? Journal of Cultural Economics 17(June), 127.
Zolberg, Vera L. (1990). Constructing a Sociology of the Arts. Cambridge: Cambridge U. Press.

You might also like