ChapterPDF Available

National wealth or national culture? A multi-country study of the factors underlying the use of celebrity endorsement in television advertising

Authors:

Abstract

A content analysis of television advertising from 25 countries reveals a wide variation in the frequency of celebrity endorsement across national markets. Further analysis confirms a negative relationship between Hofstede's individualism dimension and the use of celebrities in advertising. In contrast, country scores on Hofstede's power distance, uncertainty avoidance, and masculinity dimensions are not significantly related to the use of celebrities. Moreover, economic and advertising expenditure variables do not explain cross-national differences in the use of celebrity endorsement in advertising. The paper discusses the implications of these results, limitations of the study, and directions for future research.
©Carolus L. C. Praet 2008
1
NATIONAL WEALTH OR NATIONAL CULTURE? A MULTI-COUNTRY STUDY
OF THE FACTORS UNDERLYING THE USE OF CELEBRITY ENDORSEMENT IN
TELEVISION ADVERTISING
Carolus Praet, Otaru University of Commerce, Japan
Corresponding author:
Carolus Praet
Otaru University of Commerce
3-5-21 Midori
047-8501 Otaru
Japan
Email: praetcar@res.otaru-uc.ac.jp
Tel: +81-134-27-5349
(Fax: +81-134-27-5349)
Reference:
Praet, Carolus L.C. (2009), “National wealth or national culture? A multi-country study of the
factors underlying the use of celebrity endorsement in television advertising,” in: De
Pelsmacker, P. and N. Dens (Eds.), Research in Advertising: The Medium, the Message, and
the Context. Antwerpen: Garant, pp. 383-392.
This is the author's final pre-publication draft of the paper
ABSTRACT
A content analysis of television advertising from 25 countries reveals a wide variation in the
frequency of celebrity endorsement across national markets. Further analysis confirms a
negative relationship between Hofstede’s individualism dimension and the use of celebrities
in advertising. In contrast, country scores on Hofstede’s power distance, uncertainty
avoidance, and masculinity dimensions are not significantly related to the use of celebrities.
Moreover, economic and advertising expenditure variables do not explain cross-national
differences in the use of celebrity endorsement in advertising. The paper discusses the
implications of these results, limitations of the study, and directions for future research.
INTRODUCTION
The use of celebrity endorsers is one of the most popular advertising strategies in the United
States (Agrawal and Kamakura, 1995; Miciak and Shanklin, 1994). Miciak and Shanklin
(1994), suggest that the use of celebrity endorsement is also becoming more common in other
countries. There is however, very little empirical research on the extent of celebrity
endorsement across the world. Several academic papers provide estimates on the use of
celebrities in U.S. television advertising, ranging from more than 10% to approximately 25%
(Erdogan 1999; Erdogan et al., 2001, Walker et al., 1992). Similarly, marketing campaigns
are said to feature celebrities 20% of the time in the United Kingdom (Erdogan et al., 2001).
However, most of this information is based on industry estimates and not on the results of
empirical research. Moreover, there is a relative absence of information and research
pertaining to the use of celebrity endorsement in other countries. Exceptions include Paek
©Carolus L. C. Praet 2008
2
(2003), who reports research from Korea indicating that 32% of TV commercials, and 59% of
prime time TV commercials employ celebrities, and Kilburn (1998) who mentions that more
than 70% of Japanese television ads feature celebrities.
In addition to the relative lack of single-country empirical evidence, few studies have
compared the use of celebrities in advertising across countries. A study by Lin (1993), absent
concrete percentages, found that male celebrities were more often present in Japanese than in
U.S. television advertising, although no such difference was evident for female celebrities.
Belk and Bryce (1986) report that 41.3% of Japanese television advertising versus 11.8% of
US advertising used high status spokespeople. However, these studies generally fail to offer
explanations for the observed differences in the use of celebrity endorsement across countries.
In contrast, studies by Paek (2003) and Praet (2001) use Hofstede’s dimensions of culture to
explain differences in quantitative use of celebrity endorsement in advertising across national
markets. Paek (2003) reveals that newspaper ads in Korea featured a greater number of
celebrity endorsers (24.1%) than those in the US (9.9%) and attributed this finding to Korea’s
higher scores on Hofstede’s uncertainty avoidance and power distance dimensions of culture.
Praet (2001) introduces research findings on the use of celebrity endorsement in Japan
(47.62%), the U.S. (20.69%), France (15.57%), Germany (9.64%), Spain (6.90%), and the
Netherlands (3.92%) and suggests that differences in national scores on Hofstede’s
masculinity dimension may explain the variation in celebrity endorsement across these
countries. While the latter two studies attempt to advance the study of cross-national use of
celebrity endorsement towards offering structural explanations underlying the disparities
found, the weakness of comparing any phenomenon in only two or a few national markets
and attributing observed differences to cultural factors is that relationships may be merely due
to chance. Consequently, in order to explain dissimilarities among countries in the use of
celebrity endorsement through cultural variables and to draw more valid and generalizable
conclusions about relationships among variables, it is necessary to obtain a larger sample of
countries.
Thus, to fill the gaps in the literature as identified above, this study aims to contribute in two
ways. First, it provides additional information on the extent of celebrity endorsement across
national markets through a content analysis of television advertising from a large number of
countries. Second, it empirically tests hypotheses linking Hofstede’s dimensions of national
culture to the use of celebrity endorsement across multiple countries.
THEORETICAL BACKGROUND
Four of Hofstede’s dimensions of national culture may affect perceptions and attitudes
regarding celebrities, and consequently may affect the use of celebrity endorsement in
advertising.
Hypotheses Development
Individualism and celebrity
In contrast to people from collectivist cultures, people from individualist cultures are
generally more self-oriented and less easily influenced by others, with the exception of close
friends and nuclear family members (Hofstede, 2001). Results of a meta-analysis study by
Bond and Smith (1996) of 133 studies drawn from 17 countries examining the effects of
conformity show that the tendency to conform to others was stronger in collectivist than in
individualist countries. Celebrities can be thought of as opinion leaders and role models,
whose behavior people seek to emulate or imitate through a process of identification. Since
people in individualist cultures are less inclined to conform, one can expect consumers from
individualist cultures to be less influenced by celebrity endorsements. Conversely, one may
©Carolus L. C. Praet 2008
3
assume that in collectivist societies, consumers will exhibit a stronger inclination to perceive
celebrities as opinion leaders and role models. Further, consumers from collectivist cultures
may choose to use celebrity behavior as a social ‘template’ for their own behavior.
In addition, Hofstede (2001) reports a strong correlation between his individualism-
collectivism dimension and Hall’s (1976) typology of low-high context communication.
Gudykunst and Ting-Toomey (1988) argue that the low-high context concept and the
individualism-collectivism dimension overlap each other. In collectivist cultures, the speaker
is part of the context of the message, so the source of communication is more likely to be used
as a clue for making judgments about the meaning (Triandis, 2004). Thus, one can surmise
that the source of the advertising message in collectivist cultures will be more important than
the verbal message. The above arguments increase the likelihood of advertisers in collectivist
cultures using celebrities as endorsers. Based on the above reasoning, this research proposes
the following hypothesis:
H1: The cultural dimension of individualism is negatively related to the use of celebrities
in advertising.
Masculinity and celebrity
People in masculine cultures tend to place higher value on winning, success, and status than
people in feminine cultures do. Celebrities are usually admired, perceived as successful, and
have social status. In the United States, a masculine culture, predominant popular culture is
characterized by the cult of personality and obsession with celebrity and stardom. In feminine
cultures, for example Scandinavian countries, people frown upon the conspicuous display of
success, status, or wealth. Modesty for all, including successful people, is much more valued
as an admirable character trait in these feminine cultures (De Mooij, 2005). Consequently,
this study hypothesizes as follows:
H2: The cultural dimension of masculinity is positively related to the use of celebrities in
advertising.
Uncertainty avoidance and celebrity
People in low uncertainty avoidance cultures are relatively comfortable with ambiguity and
are tolerant of behaviors and opinions of others. Uncertainty avoidance is also relevant to the
behavior of consumers. Where uncertainty avoidance is high, consumers rely more on the
advice of people they consider experts (Hofstede, 2001). Zandpour et al. (1994) analyzed the
content of TV commercials collected from the U.S., Mexico, U.K., France, Germany, Spain,
Taiwan, and Korea. They found that the need for logical reasoning and visual information in
cultures with high uncertainty avoidance led consumers to rely more frequently on
trustworthy sources of information. Thus, consumers in high uncertainty avoidance cultures
may perceive that uncertainty is ameliorated when communication comes from the celebrity
endorser, in their minds, a credible, trustworthy, and knowledgeable source.
Hence, the following hypothesis is postulated:
H3: The cultural dimension of uncertainty avoidance is positively related to the use of
celebrities in advertising.
Power distance and celebrity
Power distance refers to social tolerance for unequal distribution of power in society and
organizations (Hofstede, 2001), or the tolerance of authority as a basic facet of one’s society
(Gudykunst et al., 1996). When compared to people from low power distance cultures, people
©Carolus L. C. Praet 2008
4
in high power distance cultures tend to obey the recommendations of public and authority
figures such as celebrities and high-status figures (Zandpour et al., 1994). In these societies,
the judgments of people who hold power are considered correct; these people function as
referent groups (Albers-Miller and Gelb, 1996). Hofstede (2001) writes that, in countries of
high power distance, this ‘referent power’ causes the less powerful to be influenced by
personal charisma of the powerful. The above reasoning leads to the following hypothesis:
H4: The cultural dimension of power distance is positively related to the use of
celebrities in advertising.
METHOD
Sample
The author and assistants recorded 7728 unduplicated television commercials in 25 countries
between February 2001 and December 2003. Television channels and days of the week were
randomly selected and rotated between the hours of 18:00 and 24:00. To ensure statistical
integrity, a sufficient sample size of commercials was collected from each country. Sample
sizes ranged from a minimum of 123 for Malaysia to a maximum of 620 for Japan. Recording
dates, time slots, and names of television channels for each country are shown in Table 1.
©Carolus L. C. Praet 2008
5
Coding Procedure
Coders received written instructions in English containing coding procedures and detailed
written explanations and definitions of all variables used in the study. Each commercial was
coded for the following variables: company name, brand name, and the use of characters,
human characters, and celebrities. Celebrities were defined as “famous people or people
whose name or face is widely known by the general population of a given country”. For each
country, two native coders independently coded a portion of the sample containing 30 to 40
commercials. Intercoder agreement was generally high among coder pairs for all countries
and across variables. A resolution procedure was used to resolve coder disagreements. Each
of the coders then coded half of the sample.
Dependent variables, independent variables and control variables
The total number of unduplicated sample commercials featuring human characters and
retained for analysis was 6359. For each of the 25 countries in the sample, a country-level
measure of the use of celebrities was constructed by dividing the number of celebrity
commercials by the total number of unduplicated commercials featuring human characters.
This measure was used as the dependent variable in the regression analysis. As independent
variables, four cultural variables, one for each of Hofstede’s four dimensions, and three
control variables were used. The control variables were Gross National Income (GNI), GNI
per capita (World Bank, 2004), and per capita television advertising expenditure (WARC,
2006). Control variables were selected for years or averages of years that match the periods of
data collection. Hofstede (2001) suggests to always routinely control for GNP per capita
when studying the influence of cultural variables on social phenomena across cultures. Since
using celebrities is more expensive than using unknown people, this suggests that advertisers
in wealthier nations may rely more on the use of celebrity endorsement.
RESULTS
Celebrity use ranged from 3.9% in Switzerland to 61.1% in South Korea. Figure 1 presents
celebrity endorsement percentages for all countries.
To test the hypotheses, an ordinary least squares regression model was used. In the full model,
the first variable, individualism is highly significant in the expected direction (β = -.770, p
= .005). The three remaining variables power distance (β = -.096, p = .697), masculinity (β
= .125, p = .562), and uncertainty avoidance (β = .078, p = .696) were not significant. A
smaller model that contains only individualism and the three control variables is also
significant for individualism (β = -.707, p = .002). Further, the adjusted R2 for this smaller
model is .333, higher than for the full model, which suggests some evidence of
multicollinearity in the full model. Regression results support H1 and do not support H2, H3,
or H4. These results and values for the economic and advertising control variables are
presented in Table 2.
DISCUSSION, LIMITATIONS AND FUTURE RESEARCH
First, this research finds that compared to other countries, the frequency of celebrity
advertising is relatively limited in continental Europe and other western countries, with the
exception of Australia, where it is relatively more common. In contrast, advertisers in Brazil,
China, Hong Kong and Malaysia use celebrities in about 25% of all advertisements, whereas
advertisers in Japan and Korean make extensive use of celebrities in almost 50% or over 60%
respectively, of all advertisements featuring human characters. This suggests that it may be
©Carolus L. C. Praet 2008
6
useful for international advertisers to distinguish three groups of countries characterized by
different levels of quantitative use of celebrities.
Second, the findings suggest that the use of celebrity endorsement may be lower in the U.S.
and the U.K. than most of the estimates and findings quoted in prior studies. A recent study
by Choi et al. (2005) that compared celebrity endorsement in U.S. and Korean television
advertising, sampled television advertising around the same period as this study. Despite
differences in methodology, (analyzing duplicated ads versus unduplicated ads in the current
study), Choi et al. report celebrity frequency percentages similar to the findings of this study:
for Korea, they report 57% versus 61% (this study); for the U.S., they report between 8.9%
and 9.6%, versus 9% (this study). This lends further support to the findings of this research, at
least for these two countries.
Third, the findings of this study suggest that Hofstede’s individualism dimension can be
useful for predicting the actual frequency of celebrity use in advertising around the world.
The more collectivist national cultures are, the more advertisers tend to use celebrities in
advertising. This knowledge, combined with a classification into the three groups of countries
identified above, may assist international advertisers in their decision making regarding the
cultural and market acceptance of using celebrities in different country markets.
Fourth, and perhaps surprisingly, the use of economic and advertising expenditure variables
does not predict the extent to which advertisers employ celebrities across national markets.
Finally, it is important to note that, contrary to the hypotheses, Hofstede’s masculinity,
uncertainty avoidance, and power distance dimensions were not significantly related to
quantitative measures of celebrity advertising. There are several possible explanations for this
finding. A first explanation is that these three cultural dimensions do not affect the
quantitative use of celebrity endorsement. A second explanation is that these dimensions may
influence the use of celebrity endorsement only at a qualitative level, which was not the focus
of this study. A third explanation is that the results may reflect that especially international
and global advertisers currently do not sufficiently understand and adjust to the cultural
differences among local consumers when using celebrities in their global or pan-regional
advertising campaigns.
Although this exploratory study featured a relatively large sample of countries, and recent
empirical studies report similar percentages of celebrity endorsement for some of the sample
countries, it is not yet possible to generalize its findings. The possibility of sampling error for
countries with small sample sizes should be considered. Sampling periods for all countries
were not perfectly matched and possible shifts in the use of celebrity endorsement over time
may have affected the findings. Future research should validate the findings of this study and
sample additional countries.
Another potential problem is that Hofstede dimension scores for some of the sample countries
in this study (Bulgaria, China, Czech Republic, Hungary, and Poland) are based on estimates
(Hofstede, 2001) and are not strictly comparable to the scores for countries included in
Hofstede’s original study. These estimated scores are a potential source of error in the
regression analyses.
Furthermore, this study did not address the influence of pan-regional advertising campaigns,
which are not designed to reflect individual country cultures, but rather reflect the culture of
the advertiser or advertising agency.
Future studies also need to address the qualitative aspects of celebrity endorsement across
cultures and should investigate the influence of product category (cf. Choi et al., 2005).
Finally, in order to derive more implications for advertisers, future research should
supplement the results of the current study by exploring perception, acceptance and liking of
celebrity endorsement advertising among consumers in different cultures.
©Carolus L. C. Praet 2008
7
REFERENCES
Agrawal, J., Kamakura, W. (1995). The Economic Worth of Celebrity Endorsers: An Event
Study Analysis, Journal of Marketing, 59(3), 56-62.
Albers-Miller, N., Gelb, B. (1996). Business Advertising Appeals as a Mirror of Cultural
Dimensions: A Study of Eleven Countries, Journal of Advertising, 25(4), 57-70.
Belk, R., Bryce, W. (1986). Materialism and Individual Determinism in U.S. and Japanese
Print and Television Advertising, Advances in Consumer Research, 13(1), 568-572.
Bond, R., Smith, P (1996). Culture and conformity: a meta-analysis of studies using Asch's
(1952b, 1956) Line Judgment Task, Psychology Bulletin, 119(1), 111-137.
Choi, S., Lee, W., Kim, H. (2005). LESSONS FROM THE RICH AND FAMOUS - A Cross-
Cultural Comparison of Celebrity Endorsement in Advertising, Journal of Advertising,
34(2), 85-98.
De Mooij, M. (2005). Global Marketing and Advertising - Understanding Cultural Paradoxes,
2nd ed., Thousand Oaks: Sage Publications.
Erdogan, B., (1999). Celebrity Endorsement: A Literature Review, Journal of Marketing
Management, 15(4), 291-314.
, B., Baker, M., Tagg S. (2001). Selecting Celebrity Endorsers: The Practitioner's
Perspective, Journal of Advertising Research, 41(3), 39-48.
Gudykunst, W., Ting-Toomey, S. (1988). Culture and Interpersonal communication,
Newbury Park: Sage Publications.
, Matsumoto, Y., Ting-Toomey, S., Nishida, T., Kim, K., Seyman, S. (1996).
The influence of cultural individualism-collectivism, self construals, and individual
values on communication styles across cultures, Human Communication Research, 22(6),
510-543.
Hall, E. (1976), Beyond Culture, New York: Anchor Books/Doubleday.
Hofstede, G. (2001), Culture’s consequences: Comparing values, behaviors, institutions, and
organizations across nations, 2nd ed., London: Sage Publications.
Kilburn, D. (1998). Star Power, Adweek, January 12, 20.
Lin, C. (1993). Cultural Differences in Message Strategies: A Comparison Between American
and Japanese TV Commercials, Journal of Advertising Research, 33(5), 40-48.
Miciak, A., Shanklin W. (1994). Choosing celebrity endorsers, Marketing Management, 3(3),
50-59.
Paek, H. (2003). Understanding Celebrity Endorsers in Cross-Cultural Contexts: An
Exploratory Analysis of South Korean and US Advertising, paper submitted to the
Association for Education in Journalism and Mass Communication 2003 Annual
Conference--Advertising Division, Kansas City, MO, retrieved from
http://list.msu.edu/cgibin/wa?A2=ind0309c&L=aejmc&F=&S=&P=1518
Praet, C. (2001). Japanese Advertising, the World's Number One Celebrity Showcase? A
Cross-Cultural Comparison of the Frequency of Celebrity Appearances in TV
Advertising, in: M. Roberts and R. King (Eds.) International Conference Proceedings of
the American Academy of Advertising, Gainesville, FL: American Academy of
Advertising, 6-13.
Triandis, H. (2004). Dimensions of culture beyond Hofstede, in: H. Vinken, J. Soeters, P.
Ester (eds.): Comparing cultures: dimensions of culture in a comparative perspective,
Brill: Leiden-Boston, 28-43.
Walker, M., Langmeyer, L., Langmeyer D. (1992). Celebrity endorsers: do you get what you
pay for?, The Journal of Services Marketing, 6(4), 35-42.
WARC (World Advertising Research Center) (2006). World Advertising Trends 2006,
Henley -on-Thames, Oxon.
©Carolus L. C. Praet 2008
8
World Bank (2004). World Development Indicators database,
http://devdata.worldbank.org/dataquery/
Zandpour, F., Campos V., Catalano, J., Chang C., Cho, Y., Hoobyar, R., Jiang, S., Lin, M.,
Madrid, S., Scheideler, H., Osborn S. (1994). Global Reach and Local Touch: Achieving
Cultural Fitness in TV Advertising, Journal of Advertising Research, 34(5), 35-63.
©Carolus L. C. Praet 2008
9
Country Recording period Time of day Source
Australia January 15-February 15, 2003
18.00-24.00
Channel 7/Channel
9/Channel 10
Austria January 15-February 15, 2003
18.00-24.00 ORF 1/ORF 2
Belgium** September 8-November 4, 2003 18.00-24.00 VTM/KA 2/VT4
Brazil September 17-November 3, 2001 18.00-24.00 TV Globo/Bandeirantes/SBT
Bulgaria September 6-October 27, 2001 18.00-24.00 Channel 1/ bTV/Nova TV
China May 23, 2001 19.30-23.00 CCTV-1
October 4-October 27, 2001 18:00-24:00 CCTV-1/CCTV-2
Czech Republic October 14-November 1, 2003 18.00-24.00 NOVA TV, Prima TV, CT 1
Finland November 13-December 18, 2001 18.00-24.00 MTV 3/Nelonen
France October 28-November 17, 2001 18.00-24.00 TF1/ France 2
Germany April 20-21, 2001 19:00-22:00 RTL
November 9-November 26, 2001 18:00-24:00
RTL/SAT1/Kabel 1/
ZDF/Pro7/RTL 2
Hong Kong May 31, June 3, 2001 20:00-23:00 TVB Jade
September 13-November 12, 2003 18.00-24.00 TVB Jade/ATV Home
Hungary October 14-November 1, 2003 18.00-24.00 RTL Klub/ TV2
Italy February 8-February 26, 2002 18.00-24.00
RAI 1/RAI 2/RAI 3/Rete 4
/Canale 5
Japan September 3-October 6, 2001 18.00-24.00 Fuji TV/TV Asahi/TBS/NTV
March 6-March 18, 2003 18.00-24.00 Fuji TV/TV Asahi/TBS
Malaysia November 8- December 3, 2001 18.00-24.00 TV 3
Netherlands March 1-7, 2001 18.00-24.00 RTL 4/SBS6/Ned 1/Ned 2
June 11-October 23, 2003
18.00-24.00 RTL 4/SBS6/Ned 1/Ned 2
New Zealand February 20, 2001 18.00-24.00 TV1/ TV2
October 10-October 29, 2001 18.00-24.00 TV1/ TV2/ TV3
Norway November 19-December 2, 2003 18.00-24.00 TV2/TV3/TV Norge
Poland October 14-November 1, 2003 18.00-24.00 Polsat/TVP 1/TVP 2
South Korea May 18-19, 2001 18:30-21:30 MBC
October 22-November 14, 2001 18.00-24.00 MBC/SBS/KBS 2
Spain March 5-May 15, 2002 18.00-24.00 La 2/Antena 3/Tele 5
Sweden September 17-October 15, 2001 18.00-24.00 TV3/TV4
Switzerland* November 5-November 10, 2001 18.00-24.00 TSR1
United Kingdom March 9, 2002 18:00-24:00 ITV
December 8-December 16, 2003
18.00-24.00 ITV/Channel 4
United States February 27-March 11, 2002 18.00-24.00 NBC/ABC/CBS/Fox
*Commercials were recorded in the French-speaking part of Switzerland
**Commercials were recorded in the Dutch-speaking part of Belgium
Table 1 Sampling period, time of day and source
©Carolus L. C. Praet 2008
10
! (Standardized)
p-Value VIF
Power Distance -.096 .697 1.894
Individualism -.770*** .005 1.863
Masculinity .125 .562 1.449
Uncertainty Avoidance .078 .696 1.237
GNI .237 .290 1.510
GNI per capita .121 .653 2.239
TV advertising expenditure per
capita
Total adjusted R2.254
Individualism -.707*** .002 1.356
GNI .265 .189 1.370
GNI per capita .087 .697 1.740
TV advertising expenditure per
capita
Total adjusted R2.333
**p < .05 (one tailed tests)
!"#$%
&"'$%
&"($%
&")$%
&"*$%
)"+$%
)"*$%
*"+$%
*")$%
,"'$%
#"-$%
#"+$%
#",$%
--"'$%
--")$%
--"#$%
-!"+$%
-!"#$%
-,"($%
(+")$%
(&"&$%
()"'$%
()"+$%
+#"'$%
)-"-$%
./012345678%
923:;%<3=>?50:%
./3837%
@A4/6B%
C35D0>E%
F>7D64B%
.=607%
G>H1406%
C>5D6406%
@3/%I365678%
J075678%
K.G%
LA5678%
K70138%M07D8AE%
J467:3%
N165B%
@31;345678H%
O34E67B%
G>H146506%
9;076%
FA7D%MA7D%
C46205%
P656BH06%
Q6=67%
.A>1;%MA436%
Figure 1 Percentages celebrity use by country, in descending order
... Celebrities play an important role in East Asian advertising. A previous study has shown that Japan and South Korea have the highest rate of celebrities in advertising with figures or more than 50% (Praet, 2009). In Japan, Hagiwara et al. (2009) showed that more than 60% of television advertisements included celebrities (62.0% in 1997; 67.4% in 2007). ...
... Previous studies have revealed that celebrities are important in East Asian advertisements with figures of 50% or more in South Korea and Japan (Choi et al., 2005;Hagiwara et al., 2009;Lee and Choi, 2012;Praet, 2009). In contrast to these findings, celebrities played only a small role among Others in all three advertising samples in which non-celebrities clearly dominated. ...
Article
Full-text available
This study conducted a content analysis of 442 television advertisements from Hong Kong, Japan, and South Korea to determine their representations of ‘Others’. Findings reveal that in East Asian advertisements, Others are overrepresented, mostly non-celebrities, and depicted in major roles, which is in contrast to previous studies in the United States where ‘Others’ are generally shown in minor or background roles. The results also reveal that Others are predominantly white, demonstrating the importance of whites in these societies and representing some type of racial/ethnic hierarchy. However, Others are also depicted as separate from the majority population. For example, they are usually depicted abroad rather than in a local setting, and they rarely interact with the local population, which sends a message of exclusion and might lead to the conclusion that they do not belong to their respective location. Possible effects and practice implications of these findings are discussed.
... Furthermore, the issues that has accounted for cultural dimensions in celebrity research have mostly concentrated on cultural orientations on either consumers or the celebrity. Again, Hofstede's individualism/collectivism dimensions have been greatly used in the celebrity and consumer behaviour studies (Baniya, 2017;Biswas, Hussain & O'Donnell, 2009;Erdogan, 1999;Jafar, Adidam, & Prasad, 2011;Park, 2019;Praet, Pelsmacker, & Dens, 2009). However, there are other forms of cultural dimensions that have received little attention which ought to be addressed. ...
Article
Full-text available
The study focuses on artefacts and values of celebrities in influencing the consumer purchase intentions. Though much is not known in this area, it enhances the discussion on the best way to increase appeal and to ensure return on investments for firms who hire the celebrities. Respondents were conveniently selected and SEM was used to carry out the analysis. The study concludes that artefacts and values have significant influence on the purchase intentions of consumers. Implications for theory and practice are discussed.
... A recent report shows that about 70-75 per cent of Korean ads include celebrities, as opposed to 10-15 per cent in most other developed countries (Turnbull, 2012). In fact, the top three countries for celebrity endorsement usage are South Korea, Japan and Malaysia (Praet et al., 2009). More interestingly, most of the time the celebrities featured in advertisements in East Asian countries have no direct connections to the products they endorse (e.g. a singer endorses a bank) and some popular celebrities can endorse numerous products. ...
Purpose The aim of this article is to summarize the celebrity endorsement literature to identify trends and challenges related to key research areas. Based on a critical review of existing literature, this article presents several recommendations regarding potential future directions of celebrity endorsement research in hospitality and tourism. Design/methodology/approach The article presents a critical review of literature from both the general marketing and hospitality and tourism fields. Findings Over the past decade, significant progress has been made in hospitality and tourism celebrity endorsement research, with several new constructs being revealed and tested. However, the extant findings are rather mixed and inconclusive because industry features have not been systematically examined and study contexts and samples have varied widely. To advance the hospitality and tourism celebrity endorsement research, an extended meaning transfer model with six propositions is proposed. Several areas for future research are also discussed. Practical implications This article offers up-to-date findings on celebrity endorsement to practitioners, and the proposed extended meaning transfer model can provide marketers useful guidelines on selecting appropriate endorsers for their products/brands. Originality/value In previous studies, scholars mainly employed one or more of the three types of celebrity endorser selection models and only examined specific antecedents of effective endorsement. To date, researchers have not yet conceptualized a modified model that captures the unique features of the hospitality and tourism industry and reconciles the mixed findings in the extant literature. This article proposes an extended meaning transfer model to explain the endorser selection process, provides a good foundational understanding of the extant celebrity endorsement research, and makes several recommendations regarding future research directions for hospitality and tourism scholars with implications for practitioners.
... In the US, according to Shimp and Andrews (2013), television and print advertisements often feature celebrities. The use of celebrity endorsement in Japan is particularly popular, as about 70 percent of all commercials feature local or foreign celebrities (Praet, 2008). Consumers often admire celebrities and trust them to be a credible source of information (Atkin & Block, 1983), which gives them power as product spokespersons, so using celebrities in advertising attracts prospective consumers and increases materialism (McCracken, 1989;Petty, Cacioppo, & Schumann, 1983). ...
Article
Full-text available
This research seeks to determine what makes young adults materialistic. The study examines the mediating role of materialism between the contextual factors and compulsive buying. Data was gathered from 219 Pakistani undergraduate university students. Partial Least Square (PLS) technique was used to analyze the data. The study confirms the intuition that more materialistic young adults are more likely to be involved in compulsive buying than are less materialistic young adults. The results were similar with the previous literature conducted in the western culture, indicating that also applies in a modern Islamic society. The findings of the study reveal that materialism mediated the relationship between certain sociological factors (i.e., group, media Celebrity endorsement, and TV advertisement) and compulsive buying. The study highlights the importance of understanding young adults' materialistic attitudes and consumption decisions and provides key knowledge for researchers, policymakers, and managers of leading brands.
... Other examples of findings with the Hofstede model are explaining differences in buying life insurance (Chui and Kwok, 2008), internet shopping (Lim et al., 2004), consumer innovativeness (Yeniurt and Townsend, 2003), international new product take-off (Tellis, Stremersch and Yin, 2003) and international growth of new products (Stremersch and Tellis, 2004). The Hofstede dimensions also have been used for comparing the use of appeals in advertising (Albers-Miller and Gelb, 1996;Chan and Moon, 2005;Emery and Tian, 2014;Rhodes and Emery, 2003), status motives (Zheng, Phelps and Hoy, 2009) or the use of celebrities in advertising (Praet, 2009). ...
Article
Full-text available
Purpose The purpose of this paper is to find consumption-related similarities and differences between the three major dimensional models of national culture, to help researchers select specific models or dimensions for their cross-cultural studies. Design/methodology/approach First, a review of the theoretical background of cultural values and three models of national culture is provided: those by Hofstede, Schwartz and GLOBE. Then these models are compared through partial correlation analysis, controlling for GNP/capita of a set of 25 relevant consumer behavior-related data with country scores of 21 dimensions of the three dimensional models. Findings Of all models several dimensions explain differences in consumer behavior. Some dimensions explain values related to specific consumer behavior domains better than others. Only a few dimensions of different models do not show meaningful interesting relationships with consumer behavior issues. Dimensions with the same label do not explain similar differences. Practical implications Cross-cultural researchers can choose from the several cultural models, but selecting a model only based on descriptions of the contents of dimensions is difficult. The relationships of dimensions with concrete consumer behavior data found in this study facilitate choice. This analysis may help researchers who consider conducting cross-cultural analysis of consumer behavior data to select a specific model, or specific dimensions of different models that apply best to their research question. Originality/value This is the first study that compares the three major dimensional models with examples of consumer behavior-related items.
Article
Full-text available
Ads often feature celebrities or others similar to the target viewer and thereby evoke envy. Envy occurs when people make an upward social comparison, and evoked envy can be either benign or malicious. The authors propose that people with different self-construals feel different degrees of benign and malicious envy depending on who is being envied: a celebrity or a similar other. Three studies were conducted comparing Americans to Koreans (Study 1), Americans to the Chinese (Study 2), and Koreans with different self-construals (Study 3). The results showed that people with high independence showed less benign envy toward the celebrity ad than toward the similar others ad, while people with low independence showed the opposite pattern. People with high interdependence showed less malicious envy toward the celebrity ad than toward the similar others ad, while people with low interdependence showed the opposite pattern.
Article
Prior literature in advertising tends to view warmth as an emotion arises from relationships between two or more characters in a commercial (e.g. lovers or family members), while little attention has been paid to endorsers as a potential source of warmth. In this article, we propose that the warmth of endorsers can serve as a way to elicit feelings of warmth in consumers and affect the effectiveness of the advertisement. Across three studies, we reveal a warmth effect in brand endorsement such that consumers perceive a greater psychological connection between the brand and the self when the brand is endorsed by a male endorser with high warmth, which, in turn, leads to a more favourable brand attitude. This warmth effect occurs only for male endorsers, but not for female endorsers. We further identify two boundary conditions and show that the effect is more prominent when consumers are of interdependent self-construal (vs. independent self-construal) and when the endorsed brand is a functional brand (vs. a prestigious brand). The findings contribute to the literature of brand endorsement, warmth and gender stereotype in advertising, and provide managerial insights for selecting brand endorsers.
Chapter
Due to digitalization and the accompanying rise of the internet over the past decades, more and more people use social media (e.g. Instagram, Facebook, YouTube) to communicate to each other. In 2017, there have been 2.46 billion users, which means that more than a third of the global population, and almost three quarters of the internet users do use at least one social media channel once a month. This trend is even going on, as the amount of social media users is expected to exceed the mark of 3 billion in 2021 (eMarketer 2017).
Article
Current theories of source persuasion and endorser credibility posit that celebrities become ineffective endorsers when they have been involved in a scandal. However, little attention has been given to the success of “bad-boy” celebrities, endorsers who thrive in spite of, if not because of, their association with negative information. In this conceptual article, we propose a new fantasy-based relationship, hedonic consumption, and congruency (FHC) model to provide insight into the effectiveness of bad-boy celebrity endorsers. According to this model, relationships with a bad-boy celebrity are formed via consumption of the celebrity’s narratives through the consumers’ use of books, magazines, movies, music, social media, television shows, or other forms of media. This view of celebrity endorsements suggests that consumers relate to bad boys through fantasy-based relationships, which allow consumers to escape from the stress of their everyday lives. Hence, a celebrity’s scandal, when congruent with the fantasy, can add to the fantasy escape and enhance the fantasy-based celebrity relationship. This fantasy-based relationship is then relived and revitalized through hedonic consumption of the endorsed product. Our model provides a new perspective on factors that influence source persuasion and insight into the effectiveness of bad-boy endorsers. Areas for future research are also proposed.
Conference Paper
The growing popularity of social media networks enables individuals to acquire large audiences of up to several million people on these platforms. Companies are starting to recognize this potential especially for young target groups, and to hire these so-called 'influencers' as endorsers in social media; these individual influences sometimes even take the place of traditional celebrities in corporate advertising. The question arises whether there is a difference in perception between these two types of endorsers and if so, which moderators influence these perceptions. In this study we explore consumers' general perception of social media influencers compared to traditional celebrities. We conduct an online survey with 590 respondents who were asked to rate 14 influencers and traditional celebrities in pairwise comparisons regarding their similarity. We apply multidimensional scaling (MDS) and find substantial differences in perceptions between the two groups. Additionally, we apply property fitting with evaluations regarding six different characteristics determining endorser effectiveness. While in total traditional celebrities are evaluated more favorably, this difference diminishes for high levels of familiarity and even reverses for perceived trustworthiness and similarity to oneself. The results indicate that marketers should carefully distinguish between influencers and traditional celebrities for endorsements, as systematic differences in perceptions between the groups are likely to have an impact on endorser effectiveness. Further research is needed to investigate which moderators (e.g. advertising type, product category) influence the perception and subsequently the endorsement effectiveness of the two types of endorsers.
Article
Full-text available
Although a number of scholars have investigated effective celebrity endorser characteristics with consumer samples using experimental methods, there is only one study by Miciak and Shanklin (1994) that explored the point of view of practitioners who are responsible for the selection of celebrities. This paper investigates British advertising agency managers' consideration of important celebrity characteristics when selecting an endorser and these factors' importance according to product types. The research findings validate much of the consumer-based research in that managers consider a range of criteria when choosing celebrity endorsers and indicate that the importance of the criteria depends on the product type.
Article
Full-text available
Use of celebrities as part of marketing communications strategy is a fairly common practice for major firms in supporting corporate or brand imagery. Firms invest significant monies in juxtaposing brands and organisations with endorser qualities such as attractiveness, likeability, and trustworthiness. They trust that these qualities operate in a transferable way, and, will generate desirable campaign outcomes. But, at times, celebrity qualities may be inappropriate, irrelevant, and undesirable. Thus, a major question is: how can companies select and retain the 'right' celebrity among many competing alternatives, and, simultaneously manage this resource, while avoiding potential pitfalls? This paper seeks to explore variables, which may be considered in any celebrity selection process by drawing together strands from various literature.
Article
Full-text available
Across cultures, do systematic differences in advertising content mirror predictable differences in the cultures themselves? The authors designed a study to shed light on that question, using Hofstede's cultural model as a tool for analyzing cultures and using advertising appeals identified by Pollay. After coding advertisements in business publications from 11 countries for the appeals employed, they computed correlation coefficients relating the proportional use of each appeal and Hofstede's cultural dimensions: individualism, uncertainty avoidance, power distance, and masculinity. The culture-reflecting quality of advertising was supported for 10 of 30 hypothesized relationships, and for an additional eight after removal of outliers from the data.
Article
Many advertising campaigns use celebrities to endorse a product but how successful is this approach? Examines the results of recent celebrity endorser studies and explores the implications for managers. Uses Madonna and Christie Brinkley to endorse bath towels, jeans and video cassette recorders; so as to assess the impact of combining endorser and product images. Concludes that the celebrity image, when combined with particular products, tends to be passed on to the product. Shows that a product which lacks a well-defined image can have one created for it through use of an endorser whose image reflects the image which an advertiser wants for that product. Conversely, using a celebrity with the wrong image or “match” will be detrimental to the image of the product.
Article
Using celebrities to promote products is a popular advertising technique around the world. However, little is known about how the implementation,of celebrity endorsement,varies according to dominant cultural values. This study content-analyzedtelevision commercials featuring celebrities from two diametrically different countries—the United States and Korea—in terms of two fundamental cultural dimensions: (1) low versus high context, and (2) individualism versus collectivism. Findings of this study suggest that the strategic use and creative executions of celebrity endorsement mirror the respective prevalent cultural orientations in the two countries, although some similarities do exist. Extensive discussion and suggestions for future research are provided. Today's American society is said to be fascinated with celebri-