Vous êtes sur la page 1sur 12

International Journal of Consumer Studies ISSN 1470-6423

Attitude towards fashion advertisements with political


content: impacts of opinion leadership and perception of
advertisement message
Beth Harben1 and Soyoung Kim2
1
Auburn University, Auburn, AL, USA
2
University of Georgia, Athens, GA, USA

Keywords Abstract
Fashion advertisements, opinion leadership,
political message, advertisement attitude, Fashion advertisements with political content are unique in that they go against the
product attitude. accepted strategy of advertising. As fashion advertisements with political content become
more prevalent, consumers’ reactions to them become an important subject of research.
Correspondence This study examines consumers’ attitudes towards fashion advertisements with political
Beth Harben, Auburn University, Auburn, AL content with respect to fashion and political opinion leadership, political views, perception
334-844-1343, USA. of advertisement message and product attitude. Results showed that attitude towards
E-mail: harbeme@auburn.edu advertisements with political content was influenced significantly by the viewer’s political
views and agreement with the advertisement message, but not by political opinion lead-
Soyoung Kim is an associate professor in ership. Fashion opinion leadership influenced product attitude only for advertisements
Merchandising at the Department of Textiles, without political content. The viewers’ advertisement attitudes were carried over to product
Merchandising and Interiors at the University attitudes regardless of the presence of political message.
of Georgia.

doi: 10.1111/j.1470-6431.2007.00648.x

The focus in a growing number of fashion advertisements in the smiling and embracing. Since then, the company’s advertisements
last 10 years has been on a political or social theme rather than on have become more controversial by removing merchandise from
the product of the company that created the advertisement. Advo- the advertisements and covering topics such as capital punish-
cacy advertising is a form of institutional advertising through ment, AIDS, world hunger, war and abortion. The advertisements
which a variety of organizations express their views on social, have featured shocking images, such as a nun and a priest kissing,
economic and political issues, and also hope to influence public a man assassinated by the Mafia and a car engulfed in flames.
opinion and behaviour (Salmon et al., 1985; Haley, 2001). An Similarly, Kenneth Cole featured in its advertising various politi-
example is the extensive campaign launched by Philip Morris in cal messages taking a stance on issues such as gun control, abor-
1996 that emphasized the dangers of youth smoking and promotes tion rights and domestic violence. In 1985, Kenneth Cole was one
the company’s youth access prevention programme. This cam- of the first apparel marketers to address AIDS awareness in
paign was part of the tobacco company’s efforts to repair its its advertisements. The advertisements displayed eye-catching
damaged public image and credibility by widely publicizing the phrases such as ‘AIDS can’t be fought on a shoe string budget,’
company’s position opposing youth smoking. While advocacy and ‘What we stand for is more important than what we stand in’
advertising is typically used to favourably position the company (Cardona, 2003). Since then, Members Only, Diesel, Sisley and
on key issues and to ultimately create a favourable public image, Express have also issued advertisements with controversial politi-
fashion advertisements incorporating political issues have differed cal messages (Kim et al., 2001). Diesel’s campaign in 1997
from the mainstream advocacy advertising in that they are aimed included violent pictures of mutilated body parts under the theme
at sparking great controversy. called ‘Diplomacy is a kick in the face’ (Andersson et al., 2004).
The Italian apparel company Benetton brought advocacy adver- The political content of fashion advertisements was quickly
tising to fashion in 1984 with its campaign, ‘All the Colors of the noticed and widely discussed by the public. For example, the
World’ (Tinic, 1997). The purpose of the campaign was to promote campaign entitled ‘We, on Death Row’, which featured pictures
racial unity, and its advertisements featured young adults from and interviews with death row inmates, caused Sears, Roebuck and
different racial groups wearing brightly coloured clothing while Company to pull Benetton US products from store shelves
(Reuters, 2000; Taylor, 2000). The campaign was condemned for
Beth Harben completed her Masters degree in Textiles, Merchandising, humanizing the inmates, and the manner in which Benetton gained
and Interiors at the University of Georgia in 2004, and is currently a PhD access to the inmates was investigated (Davis, 2001). More
student in the Department of Consumer Affairs at Auburn University. recently, Kenneth Cole launched an advertisement that negatively

88 International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd
B. Harben and S. Kim Fashion advertisements with political content

portrayed the Bush–Cheney administration. However, after the (hereafter ‘advertisement attitude’) will affect their attitudes
terrorism of September 11, the designer ended the campaign due to towards the products. Often practitioners of advocacy advertising
complaints that the advertisements were anti-patriotic. believe that their campaigns enhance their credibility and influ-
The complaints against these advertisements are not geographi- ence public opinion, but little research has been conducted to
cally restrained. During the Gulf War in 1991, Benetton created an prove the effectiveness of their campaigns (Burgoon et al.,
advertisement that contained a picture of a war cemetery. With the 1995). The current study examines consumers’ attitudes towards
exception of one Italian newspaper, mass media around the world a fashion advertisement with political content in terms of what
refused to print the advertisement. Hundreds of journalists criti- factors influence them and how advertisement attitude affects
cized the company for attempting to profit off an image of death. attitude towards the product featured in the advertisement. This
In response to the criticism, Benetton created an advertisement study has chosen to take the positivistic-empirical approach to
with an image of birth. The image of a newborn baby still attached understanding these relationships, and designed a case study
to the umbilical cord was highly condemned in Italy, France, Great where two fictitious advertisements were tested using college
Britain and Ireland. However, the advertisement eventually won students as a research group. This study also includes a deter-
awards in Sweden and the Netherlands (Benetton, n.d.). In mination of whether opinion leadership is related to advertise-
Germany, Benetton licensees sued the company, claiming that the ment attitude and product attitude. Opinion leaders are the
controversial advertisements prevented customers from wanting to people in society most affected by formal media and they
enter their stores (Nash, 1995). Although the Benetton advertise- communicate their views most frequently and more influentially
ment which featured a man dying from AIDS was praised by a than others. Therefore, their attitudes towards advertisements
German AIDS group, both Germany and France prohibited the with a political message bear potential significance for
advertisement from being publicly displayed (Tinic, 1997). marketers.
In Sweden, some of the advertisements by Diesel and Sisley It should be noted that this study examined advertisement atti-
have been reported to The Market Ethical Board, and Benetton’s tude from an American perspective, inasmuch as the political
campaigns have been reported to the Department of Justice message chosen referred to the US government’s decision to go to
(Andersson et al., 2004). war with Iraq. In the late spring of 2004, the war in Iraq was a
Institutional advertising with weak persuasiveness and incon- highly controversial topic for the presidential campaign. The issue
sistency with the corporation’s actual practices can also present sparked debate around the country; it revived an interest in politics
the danger of backfiring. For example, Shell Oil has waged a green for a historically apathetic demographic group in America, the
marketing campaign proclaiming its commitment to environmen- college student. In 1972, immediately after the government
tal issues, even while it is widely known that it is fighting lawsuits changed the voting age from 21 to 18, the percentage of people
and criticisms on a large number of fronts throughout the world for aged 18–24 who voted was 49.6% (U.S. Census Bureau, 2000).
its purportedly environmentally and socially irresponsible actions. From 1972 to 2000, the percentage of voters in this age group has
This green marketing campaign triggered environmentalists ranged around 20–30%, with a low of 16.6% in 1998 and a peak
and human rights supporters, including Greenpeace, to launch a of 42.8% in 1992 (U.S. Census Bureau, 2000, 2002). Students
massive counter-campaign in the media, slashing sales and result- have cited apathy by way of explanation pointing out that the
ing in considerable damage to the corporate public image (Doyle, politicians were not addressing issues that were important to them
2004). (Speckman, 2004). However, students believed that the war was an
Therefore, while fashion advertisements with political content important issue, as many had friends who were going over to Iraq
are unique, going against the accepted strategy of advertising, it to fight, and that the war was having a negative effect on jobs and
remains to be seen whether they will prove to be effective in the the economy (Ebenkamp, 2004). As the war in Iraq inspired debate
long run. Conventional advertising shows the positive character- and protests around many of the nation’s campuses, college stu-
istics of a product or the company itself, with the intent of selling dents were as divided as the rest of the nation. Speakers were
the product to the viewer. Many advertisements even go a step jeered off the stage for condemning the Bush administration’s
further and try to sell not only the product but also a pleasant policies on one campus and asked to leave the stage for praising
situation or a desirable quality that could result from the product’s the administrations’ policies on another campus (Marcus, 2003).
purchase. In contrast, fashion advertisements with political Therefore, this particular political issue makes it an appropriate
content differ, because they do not mention the product’s qualities. choice for the study, as it may be less likely to be subject to college
Often the product is not even pictured in the advertisements. These students’ political apathy.
advertisements may show or describe the harsh reality of the world
instead of the pleasantries of a fantasy land. Benetton claims that
these advertisements are innovative and thus spark interest, as they
Literature review
are different from what society is accustomed to experiencing. As
Fashion and political opinion leadership
Huang (1997) argues, even advertisements that elicit negative
feeling can still be favoured by consumers, thereby increasing Opinion leaders are those who exert ‘influence on the opinions of
advertising effectiveness. others within a specific field through personal communication’
As fashion advertisements with political content become more (Hellevik and Bjorklund, 1991), and who are interested in that
prevalent, their effects on consumers become an increasingly field (Weimann, 1994). Because opinion leaders in one field
important subject of research. Marketers know that these adver- are not necessarily opinion leaders in another (Hellevik and
tisements will attract consumers’ attention, but they also need to Bjorklund, 1991), we included in our study two types of opinion
know how the consumers’ attitudes towards these advertisements leadership: fashion and political opinion leadership.

International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd 89
Fashion advertisements with political content B. Harben and S. Kim

In his study of French female fashion opinion leaders, Vernette versy of the issue, but also on personal characteristics such as
(2004) states that fashion media should target opinion leaders to be agreement with the message or interest in the topic.
successful, because opinion leaders tend to be more positive Lord et al. (1995) reported that message agreement strongly
towards media advertising than non-leaders. Given that opinion influences advertisement attitude. This finding was corroborated
leadership is linked to use of the media for information seeking by Laczniak et al. (1999), who studied the influence advertising
and opinion leaders influence the attitudes of other individuals by message involvement, product involvement and product knowl-
providing a primary word-of-mouth source of information (Herr edge have on the way a consumer processes an advertisement. Of
et al., 1991; Shah and Scheufele, 2006), his argument appears well the three variables, advertising message involvement had the
substantiated. Vernette further explains that because fashion strongest influence on advertisement processing. In fact, product
opinion leaders have enduring involvement with fashion products, involvement and product knowledge had little impact on a con-
and also because advertising is their frequent topic for discussion, sumer’s response to the advertisement after controlling for the
their views carry more weight in communication with others, influence of advertising message involvement.
thereby strengthening advertising persuasion. In his study, Additionally, recognition and understanding of an advertise-
however, the respondents were asked about their attitudes towards ment message may also be a factor in determining advertisement
fashion advertising in general rather than about their attitudes attitude. For example, one study examining the effectiveness of
towards any specific advertisement. Advertisement attitude Mobil Oil’s advocacy campaign (Adkins, 1978) found that only
was measured by such questions as ‘Life would be dull without 9% of those polled understood what the advertisements were
advertising’ and ‘I find advertising entertaining’. An examination saying. The fact that more than half of the sample indicated a
of whether fashion opinion leaders will be in favour of fashion negative attitude towards the campaign may be associated with the
advertisements despite controversial political messages contained fact that the message was not understood by the majority. All these
therein should provide interesting insights into the role of fashion studies suggest that how one perceives the message in the adver-
opinion leadership in advertising effectiveness. tisement may influence one’s advertisement attitude.
As our study involved advertisements with political content, it Hypothesis 3: Perception of advertisement message will have
seemed appropriate to investigate the influence of political opinion a significant impact on attitude towards advertisements with a
leadership as another form of opinion leadership. Political opinion political message.
leaders are those who have a major influence on others’ opinions
about various political issues. A few dominant characteristics of Advertisement attitude and product attitude
political opinion leaders identified in previous studies include a
Consumers’ advertisement attitudes ultimately lead to the forma-
high level of media exposure, strong interest in political issues and
tion of attitudes towards the products featured in the advertise-
active involvement in political affairs and organizations. Hellevik
ments, because consumers use advertisement claims to make
and Bjorklund (1991) also discovered that political opinion
evaluations about product benefits (Burton and Lichtenstein, 1988;
leaders tended to hold more extreme views than non-leaders.
Kim et al., 2002; Mostafa, 2005). Although political messages in
Given that political opinion leaders are strongly interested in
fashion advertisements have little to do with endorsing products or
political topics, use mass media frequently and tend to form
promoting product attributes, it would be of interest to see whether
extreme views, we hypothesized that political opinion leadership
a close connection between advertisement attitude and product
may have a measurable impact on attitude towards advertisements
attitude could be established for this type of advertisements.
with political content.
Hypothesis 4: Advertisement attitude will have a significant
Hypothesis 1: Fashion opinion leadership will have a signifi-
impact on attitudes towards the products featured in
cant impact on attitude towards advertisements featuring
advertisements.
fashion products.
Hypothesis 2: Political opinion leadership will have a signifi-
cant impact on attitude towards advertisements with a politi-
Method
cal message.
Description of the survey
Four web pages were designed to collect data for the study. Each
Perception of advertisement message
web page included one of four jean advertisements created by the
Advertisement attitude is ‘the predisposition to respond in a researchers, as well as a link to a five-part online survey. The four
favourable or unfavourable manner to a particular advertising advertisements designed specifically for this study comprised: (1)
stimulus during a particular exposure occasion’ (MacKenzie et al., a pair of female jeans and a political message (Fig. 1); (2) a pair of
1986, p. 130). Advertisement attitude is based on one’s reactions female jeans and a non-political message (Fig. 2); (3) a pair of
to various elements of the advertisement, such as model, celebrity male jeans and a political message (Fig. 3); and (4) a pair of male
endorser, colour, layout and text message (Kim et al., 2002). As jeans and a non-political message (Fig. 4). Jeans were chosen as
Schlossberg (1992) noted, advertisements with positive social the fashion product in the advertisement, because they are com-
messages bring about positive responses by shaping public atti- monly worn by both male and female students and thus familiar to
tudes in the intended manner; however, in the context of adver- all students in the sample. A contemporary style of jeans was
tisements dealing with controversial political issues, the impact of chosen, but the jeans did not have trendy features that might
a message on consumer attitude is not clear. As illustrated by distract the respondents.
public reactions to campaigns by Benetton, Kenneth Cole and The topics that were used in the message for the advertisements
others, the impact may depend not only on the degree of contro- were also familiar to the sample. The political message was ‘While

90 International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd
B. Harben and S. Kim Fashion advertisements with political content

Figure 1 Advertisement with female jeans and a political message. Figure 2 Advertisement with female jeans and a non-political message.

the President was busy creating a reason for war, we were busy
creating the perfect pair of jeans’. The message was referring to influence people’s opinions about fashion’. The reliability index
the controversy surrounding the President’s decision to go to war for the nine-item scale was low at 0.60. Examination of item-
with Iraq. This topic was chosen because it was widely being total statistics indicated that the reliability would improve by the
discussed among college students at the time of this survey. The largest amount upon removing two questions (‘I don’t need
non-political message was ‘We have a fit for everybody in to talk to others before I buy clothing’ and ‘I like to get
America’. This message simply advertises the jeans. Aside from other’s opinions before I buy clothing’). The reliability of the
the differences in message, the advertisements were made to look remaining questions was acceptable with the Cronbach’s alpha
identical. The same background was used for both advertisements, of 0.78.
and the jeans and the message were placed in the exact same
position in the advertisements.
Political opinion leadership
Eleven questions measuring political opinion leadership were
Fashion opinion leadership
either adapted from Hellevik and Bjorklund (1991) or created
The first section of the survey included nine questions measuring by the researchers. Hellevik and Bjorklund divided political
fashion opinion leadership on a 7-point Likert scale (1 = strongly opinion leadership into three parts based on different aspects of
disagree, 7 = strongly agree). The questions were adopted from a political discussion: (1) whether the respondent speaks his or her
scale of product-specific opinion leadership developed by Flynn views during political conversation or listens to others’ views
et al. (1996). Examples of the questions included ‘Other people (1 = strongly disagree, 7 = strongly agree); (2) whether the
come to me for advice about choosing clothing’ and ‘I often respondent engages in political discussion at all (1 = never,

International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd 91
Fashion advertisements with political content B. Harben and S. Kim

Figure 3 Advertisement with male jeans and a political message. Figure 4 Advertisement with male jeans and a non-political message.

7 = daily); and (3) whether others are influenced by the respon- (‘I have strong feelings about the political message in the adver-
dent’s views (1 = strongly disagree, 7 = strongly agree). The tisement’). The respondents then proceeded to answer four ques-
researchers added five questions asking how frequently the tions regarding advertisement attitude (interesting, effective,
respondent engaged in various political activities and how fre- good, likable). The reliability for the advertisement attitude scale
quently he or she acquires political information from the media was 0.87.
(1 = never, 7 = weekly or daily depending on question). The reli-
ability for this scale was 0.81. Product attitude
Three questions adapted from studies by DeLong et al. (2002) and
Perception of advertisement message and Kim et al. (2002) were used to measure product attitude (‘The
advertisement attitude product in this advertisement is stylish’, ‘The product in this
The respondents were then asked to review the assigned adver- advertisement is attractive’ and ‘Overall, I like the product in the
tisement and asked questions regarding their (1) recognition of advertisement’) (1 = strongly disagree, 7 = strongly agree). The
the political message in the advertisement (‘I believe the adver- reliability for this scale was 0.84.
tisement is communicating a political message’); (2) understand-
ing of the message (‘I understand what the political message in
Survey participants
the advertisement is trying to say’); (3) strength of agreement
with the message (‘I agree with the political message in the In order to provide a broad sampling of college students, respon-
advertisement’); and (4) strength of feelings about the message dents were recruited from diverse fields of study (biology,

92 International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd
B. Harben and S. Kim Fashion advertisements with political content

advertising, political science, psychology, leisure studies, (n = 126, 70%). Participation by male students (n = 52, 29%) may
housing). The survey web site address was distributed to a total of have been low as fashion is still largely viewed as a feminine topic
1268 students, of whom 179 responded. One of the researchers of interest. Almost all of the students were single (n = 171, 96%).
visited each of 12 selected classrooms, briefly explained the When asked about their political views (1 = very conservative;
survey and handed out a small slip of paper with the following 4 = neutral; 7 = very liberal), the respondents tended to be neutral
information printed on it: (1) the title of the survey, ‘College with a mean of 4.32 (SD = 1.75).
Students’ Opinions of Fashion Advertisements’; (2) the web The political message group strongly believed that the adver-
address for the survey; and (3) the phrase ‘Your opinion counts!’ tisement was communicating a political message (mean = 6.23,
in order to better motivate the students to participate in the survey. SD = 1.14), and strongly indicated that they understood the
An equal number of flyers were distributed for the advertisement message (mean = 6.28, SD = 1.05). However, their overall agree-
containing a political message and for the advertisement with a ment with the message was relatively weak with a mean of 3.05
non-political message. The web sites were available for 5 weeks. (SD = 2.08). They had somewhat strong feelings about the
The low response rate of 14% may be due to the particular message (mean = 5.70, SD = 1.26), but they were not asked if their
method used to recruit participants. Each of the classes was visited feelings were positive or negative.
only once; thus there was no repeat contact with any of the stu-
dents. In addition, no incentive was offered for taking the survey.
This method of recruitment, while promoting strictly voluntary Group comparisons
participation, raised a concern as to the representativeness of the
T-tests were conducted to compare the political message group
final sample. The response rate might have been improved, if a
with the non-political message group in terms of advertisement
persuasive cover letter had been given to the students when they
attitude and product attitude (see Table 1). When summated scales
were initially contacted. A cover letter was made available only
for both advertisement attitude and product attitude were used, the
when a student correctly accessed the web page. The weak
political message group did not significantly differ from the other
response rate can be also attributed to the fact that male students
group in either variable. Additional t-tests were performed to
might not have found the subject of fashion advertisement impor-
determine whether the two groups differed for each of the
tant and intriguing enough to participate in the survey. As classes
individual items constituting advertisement attitude and product
with a balanced representation of both genders were deliberately
attitude. Significant differences were detected for two items
chosen, low participation by male students may be seen as a major
measuring advertisement attitude; the political message group
cause for the poor response rate.
believed more strongly that the advertisement they viewed was
After 5 weeks, a total of 177 usable data were generated. As the
interesting (P < 0.001), and the non-political message group
survey was completed in a volunteer manner, an even distribution
believed more strongly that the advertisement they viewed was
of respondents for each advertisement could not be ensured. Even
good (P < 0.05).
so, our results yielded an almost even distribution between the
advertisements; 84 students (47%) viewed the advertisement with
political message (hereafter to be called the political message Multiple regression analyses
group) and 93 students (53%) viewed the advertisement with no
political message (the non-political message group). Ordinary least squares regression was used to examine how
strongly advertisement attitude and product attitude were pre-
dicted by demographic characteristics and the selected perceptual
Results variables. Using the enter method of multiple regression, a set of
demographic variables was first entered into the equation followed
Description of respondents
by fashion opinion leadership, political opinion leadership and
The ages of the respondents ranged from 18 to 27 years with an political views simultaneously, then by perceptions of advertise-
average of 21 years; the majority of the respondents were female ments, and finally by advertisement attitude. In order to compare

Table 1 Comparisons between political


Political message Non-political message
message group and non-political message
(n = 84) (n = 93)
group
Mean SD Mean SD T-value

Ad attitude 3.42 1.67 3.51 1.42 -0.37


The ad is interesting 4.62 1.93 3.49 1.59 4.21***
The ad is effective 3.27 1.98 3.66 1.68 -1.39
The ad is good 2.90 1.93 3.49 1.57 -2.24*
I like this ad 2.89 1.99 3.39 1.61 -1.82
Product attitude 5.10 1.25 5.01 1.16 0.49
The product is stylish 5.26 1.43 4.89 1.45 1.70
The product is attractive 5.18 1.34 5.01 1.22 0.87
Overall I like the product 4.87 1.66 5.14 1.14 -1.25

*P ⱕ 0.05, ***P ⱕ 0.001.

International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd 93
Fashion advertisements with political content B. Harben and S. Kim

the political message group with the non-political message group, advertisement than younger and male respondents. The same two
two separate sets of regression analyses were conducted. variables, however, emerged as insignificant when other percep-
tual variables were entered into the equation. Of the perceptual
variables, only advertisement attitude was significant (b = 0.51,
Advertisement with political content
P < 0.001). The final model explained 32% of the variation in
For the political message group, the following independent vari- product attitude.
ables were included in the regression model: demographic charac-
teristics (age and gender), fashion opinion leadership, political
Advertisement without political content
opinion leadership, political views and four message-perception
variables (recognition of the message, understanding of the For the non-political message group, the two demographic vari-
message, agreement with the message and feelings about the ables, two opinion leadership variables and political views were
message) (see Table 2). The enter method was employed to examined for their impacts on advertisement attitude (see Table 2).
compare the influences of demographic characteristics and percep- Age was the only significant predictor of advertisement attitude,
tual variables. with older students being more positive towards the advertisement
The results showed that neither of the two demographic vari- (b = 0.39, P < 0.001). The R2 coefficient of the final model was
ables significantly influenced advertisement attitude. The R2 coef- 0.20.
ficient was rather low (6%) with only two variables included in the A separate regression model tested the impacts of the same
equation. When two opinion leadership variables and political variables and advertisement attitude on product attitude (see
views were added to the model, the R2 coefficient increased to Table 3). Age and gender were both insignificant when entered
0.22. Only one variable, political views, was significant (b = 0.43, alone, but emerged as significant when three other variables
P < 0.001) revealing that the more liberal the student is, the more (fashion opinion leadership, political opinion leadership, political
positive attitude he or she holds towards the advertisement. views) were added to the equation. Older (b = 0.24, P < 0.05) and
Neither fashion nor political opinion leadership was significant female students (b = 0.32, P < 0.05) held a more positive attitude
in predicting advertisement attitude. When the four message- towards the product in the advertisement. The respondent’s
perception variables were entered into the equation, the R2 coeffi- political view did not have a significant impact on product atti-
cient increased to 0.48. Of the four message-perception variables, tude. The R2 coefficient increased from 0.05 to 0.14 when the
only agreement with the message significantly influenced adver- three perceptual variables were added. When advertisement atti-
tisement attitude. Respondents who more strongly agreed with the tude was added to the equation, the R2 coefficient increased to
message exhibited a more favourable advertisement attitude 0.20. Age became insignificant but gender remained significant
(b = 0.69, P < 0.001). (b = 0.32, P < 0.05). Interestingly, both fashion and political
The second regression model tested the impacts of the same opinion leaderships emerged as significant predictors of product
variables and advertisement attitude on product attitude (see attitude; higher levels of fashion opinion leadership (b = -0.32,
Table 3). Age (b = 0.25, P < 0.05) and gender (b = 0.27; P < 0.05) P < 0.05) and political opinion leadership (b = -0.25, P < 0.05)
were significant predictors of product attitude when only demo- were associated with less favourable product attitudes. As
graphic variables were included; older and female respondents expected, advertisement attitude was a significant predictor of
responded more favourably towards the product featured in the product attitude (b = 0.28, P < 0.05).

Table 2 Regression analysis results for ad


Ad attitude
attitude
Non-political
Political message group message group

Beta Beta Beta Beta Beta

Demographic characteristics
Age 0.16 0.09 -0.03 0.41*** 0.39***
Gender (1 = female) 0.21 0.13 0.06 0.09 -0.03
Perceptual variables
Fashion opinion leadership 0.02 0.02 0.24*
Political opinion leadership 0.13 0.03 0.15***
Political views 0.43*** -0.02 -0.07
Perception of ad message
Recognition -0.25
Understanding 0.28
Agreement 0.69***
Feelings 0.05
R2 0.06 0.22 0.48 0.15 0.20

*P ⱕ 0.05, ***P ⱕ 0.001.

94 International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd
B. Harben and S. Kim Fashion advertisements with political content

Table 3 Regression analysis results for pro-


Product attitude
duct attitude
Non-political
Political message group message group

Beta Beta Beta Beta Beta Beta Beta

Demographic characteristics
Age 0.27* 0.25* 0.19 0.21 0.19 0.24* 0.13
Gender (1 = female) 0.24* 0.27* 0.23 0.20 0.18 0.32* 0.32*
Perceptual variables
Fashion opinion leadership -0.10 -0.12 -0.13 -0.25* -0.32*
Political opinion leadership 0.01 -0.12 -0.14 -0.21* -0.25*
Political views 0.22 -0.09 -0.08 -0.13 -0.11
Perception of ad message
Recognition -0.20 -0.07
Understanding 0.14 0.00
Agreement 0.20 -0.16
Feelings 0.20 0.17
Ad attitude 0.51*** 0.28*
R2 0.11 0.12 0.18 0.32 0.05 0.14 0.20

*P ⱕ 0.05, ***P ⱕ 0.001.

not place as much importance on keeping up with current events in


Discussion the world as older adults do. A larger sample with a better repre-
sentation of the general population might reveal different findings.
Hypothesis testing
Interestingly, political opinion leadership negatively impacted
on product attitude for the advertisement without political content.
Fashion opinion leadership
Although the relationship between the two variables was not pro-
Fashion opinion leadership did not have a significant impact on posed for the advertisement without political content, the finding
either advertisement attitude or product attitude for the advertise- was not completely unexpected. The negative direction of the
ment containing political content, but did significantly influence relationship may suggest a lack of interest in fashion products
product attitude for the advertisement without political content, among political opinion leaders.
offering partial support for Hypothesis 1. Interestingly, those
respondents with a higher level of fashion opinion leadership
Political message vs. non-political message
exhibited less favourable attitudes towards the product in the
advertisement without political content. This finding may be due The result of t-tests showed that the presence of a political
to the fact that the product chosen for the advertisement stimuli message did not necessarily affect the respondent’s overall adver-
was jeans, which, although contemporary in style, lacked trendy tisement attitude. This finding, however, should not be considered
features. As fashion opinion leaders tend to actively monitor the conclusive. When individual items measuring advertisement atti-
latest fashion trends (Summers, 1970; Vernette, 2004), they may tude were examined, those who viewed the advertisement with
have found the ordinary jeans less than exciting. In order to gain a political content agreed more strongly than those who viewed the
full understanding of the role of fashion opinion leadership in the advertisement without political content that their advertisement
formation of attitudes towards fashion advertisements, future was interesting, but agreed less strongly that the advertisement
studies need to include a variety of products with a varying degree was good. When individual items measuring advertisement atti-
of fashionability. tude were tested for their impacts on product attitude, ‘good adver-
tisement’ exhibited a stronger positive relationship with product
attitude (b = 0.50, P < 0.001) than ‘interesting advertisement’
Political opinion leadership
(b = 0.27, P < 0.05). These findings suggest that some advertise-
The results from multiple regression analyses also showed that ments with political content, even when perceived as interesting,
political opinion leadership did not have a significant impact on can still be considered poor, resulting in less favourable product
either advertisement attitude or product attitude for the advertise- attitudes. These findings lend support for an early study by
ment with political content, contradicting our Hypothesis 2. These Berlyne (1960), who suggested that degree of interest should not
findings, however, should not be generalized to the population at be combined into one global measure of overall advertisement
large without caution. Many of the questions measuring political attitude, as there exists a non-monotonic relationship between
opinion leadership required the respondents to indicate how often degree of interest and overall liking for an advertisement.
they use the media to gain information on current political affairs. Factors that contribute to the perception of interesting but poor
These questions may not have been as appropriate for college advertisements with political content may include the viewer’s
students as they might be for others, as many college students may political view and his or her level of agreement with the political

International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd 95
Fashion advertisements with political content B. Harben and S. Kim

message conveyed in the advertisement. As findings from multiple more favourably towards the jeans featured in the advertisements
regression analyses revealed, attitudes towards the advertisement than younger and male students. As with fashion opinion leaders,
with a political message were significantly influenced by the younger students might have found the basic-style jeans not trendy
respondents’ political views, and also by their strength of agree- enough to consider interesting. The gender difference, on the other
ment with the message in the advertisement. As the political hand, may reflect the fact that women are generally more inter-
message chosen for this study took a stance against the Republican ested in fashion items and therefore more favourable towards
President’s decision to go to war with Iraq, it was not surprising fashion advertisements than men.
that the more politically conservative the viewer was, the less
positive attitude he or she held towards the advertisement. In
addition, the less strongly the viewer agreed with the message in
Conclusions
the advertisement, the less positive attitude he or she held towards This study marks a unique effort to investigate attitude towards
the advertisement. The fact that a respondent’s political views and fashion advertisements containing political content with respect to
agreement with a message were significant predictors of adver- the effects thereon of fashion and political opinion leadership,
tisement attitude suggest that marketers should be aware of the political views and perception of advertisement message. Political
negative impact of their advocacy advertising for those who do not consumerism, however, is a complex issue that warrants consid-
share the same political views. erations of social, economic and political contexts, and therefore,
a case study involving a small sample of college students should
be reviewed with caution. Stolle et al. (2005) defined political
Agreement with advertisement message and overall
consumerism as ‘consumer choice of producers and products
advertisement attitude
based on political or ethical considerations’. Although political
Hypothesis 3 was only partially supported, as just one of the four consumerism is typically discussed in conjunction with consumer
variables measuring perceptions of advertisement message (i.e. boycotts, this study of consumers’ attitudes towards fashion adver-
agreement with the message) appeared as a significant predictor of tisements with political content can also be seen to be in the vein
advertisement attitude. Neither recognition nor understanding of of political consumerism, given that advertisement attitude often
the message affected the respondents’ attitudes towards the adver- leads to product choice. A subsequent question for discussion is
tisement. Further, emotional reactions to the message, measured whether the degree of a consumer’s involvement in his or her
by feelings towards the message, did not influence advertisement decision-making process is a prerequisite for political consump-
attitude. This finding suggests the need for further investigation tion. Will it still be considered political consumption even if the
into the roles of emotional vs. cognitive reactions to advertisement consumer is not deliberately making the decision to express his or
messages. One possible explanation from a methodological point her political opinion? The issue becomes even more complicated
of view is that the respondents might have found it easier to when an argument is made that some people may decide against
indicate their level of agreement with the message, but were not purchasing a product, not because they do not agree with the
sure of how to respond in terms of the strength of their feelings. political views contained in the advertisement, but simply because
Additionally, it is possible that the relationship between emotional the advertisement deals with a political issue. A myriad of envi-
reactions to the message and overall advertisement attitude may be ronmental and individual factors affecting a consumer’s political
significant only for either positive or negative feelings; thus, future consumption presents significant challenges to understanding the
studies need to consider the nature of the viewer’s emotional issue of advertisement with political content.
reactions with a question specifically asking whether their feelings This study took the positivistic-empirical approach and investi-
are positive or negative. gated relationships among a few selected variables in an artificial
environment. As such, it will face the usual criticisms of the
positivistic philosophy, one of which is that this mode of inquiry
Advertisement attitude and product attitude
tries to explain the rich complexity of human interaction in
Consistent with Hypothesis 4, advertisement attitude significantly simplistic patterns. Although this study suffered from over-
influenced product attitude for both types of advertisements. This simplification, it on the other hand attempted a systematic inves-
finding suggests an important implication for fashion advertisers. tigation into the effects of individual characteristics, such as
The viewer’s positive or negative attitude towards the advertise- opinion leadership, political views, agreement with political
ment can be carried over to his or her attitude towards the product, message and demographics.
even when the advertisement does not contain any message per- Although social context and group characteristics are not part of
taining to the product. If the viewer does not agree with the these analyses, some of the findings can be discussed taking these
message in the advertisement, he or she will likely respond factors into consideration. During the time of the study, the nation
unfavourably towards the advertisement, which will, in turn, lead was divided over the president’s decision to go to war in Iraq. The
to an unfavourable attitude towards the product. debate spread onto college campuses across the country. College
students, a group that had been apathetic towards politics in the
past, were now voicing their opinions. The degree of interest and
Demographic characteristics
division of opinions rendered the issue appropriate for this study.
Although age and gender effects were not a focus of this study, Leading up to the war in Iraq, the president gave a number of
regression analyses demonstrated that product attitude for both reasons to go to war, including Saddam Hussein’s cruel and inhu-
advertisement stimuli were significantly influenced by age and mane practices, his disregard for UN weapons inspections and
gender. Specifically, older students and female students responded his stockpiling of weapons of mass destruction (Purdum, 2002;

96 International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd
B. Harben and S. Kim Fashion advertisements with political content

Gordon, 2003). While the first two reasons gave cause for the Burgoon, M., Pfau, M. & Birk, T.S. (1995) An inoculation theory expla-
punishment of Hussein, many believed that this did not warrant nation for the effects of corporate issue/advocacy campaigns. Commu-
going to war. When officials found no weapons of mass destruc- nication Research, 22, 485–505.
tion, people began criticizing the president and many believed that Burton, S. & Lichtenstein, D.R. (1988) The effect of ad claims and ad
context on attitude toward the advertisement. Journal of Advertising,
he had fabricated the intelligence to justify the invasion of Iraq
17, 3–11.
(Dean, 2003). US citizens and people from all over the world Cardona, M.M. (2003) Fashion and apparel. In The Advertising Age
began to question the Bush administration’s prewar intelligence, Encyclopedia of Advertising (ed. by J. McDonough & K. Egolf),
resulting in widespread mistrust of the US and a rapid deteriora- pp. 567–571. Fitzroy Dearborn, New York.
tion of public opinion thereof. As this is a rather extreme circum- Davis, C. (2001) A killer campaign. Columbia Journalism Review, 39,
stance, it is possible that choice of a less divisive political issue 9.
for inclusion in the study might well have resulted in different DeLong, M., LaBat, K., Nelson, N., Koh, A. & Kim, Y. (2002) Global
outcomes. products, global markets: jeans in Korea and the United States. Cloth-
The findings of this study seem to reveal college students’ ing and Textiles Research Journal, 20, 238–245.
former reputation for apathy towards politics, and simultaneously, Dean, J.W. (2003) Is lying about the reason for a war an impeachable
offense? [WWW document]. URL http://www.cnn.com/2003/LAW/
their new political awareness. Students’ lack of enthusiasm for
06/06/findlaw.analysis.dean.wmd/ (accessed on 1 June 2007).
political consumerism may partially explain why political opinion Doyle, J. (2004) Rising the Dragon: Royal Dutch Shell and the Fossil
leadership had no significant impact on advertisement attitude and Fire. Common Courage Press, Monroe, ME.
product attitude for the advertisement with political content. Other Ebenkamp, B. (2004) The young and apathetic? Brandweek, 45, 18.
results, however, reveal some degree of concern among college Flynn, L.R., Goldsmith, R.E. & Eastman, J.K. (1996) Opinion leaders
students for the particular political issue chosen for this particular and opinion seekers: two new measurement scales. Journal of the
study. Students who expressed that the advertisement with politi- Academy of Marketing Science, 24, 137–147.
cal content was interesting and those who had somewhat strong Gordon, M.R. (2003) Bush enlarges case for war by linking Iraq with
feelings about the message were found less favourable towards Terrorists. The New York Times, 29 January, A1.
the advertisement. In addition, the more politically conservative Haley, E. (2001) Exploring the construct of organization as source: con-
sumers’ understandings of organizational sponsorship of advocacy
the viewer, the less positive the attitude he or she had towards the
advertising. Journal of Advertising, 24, 19–35.
advertisement. As previously pointed out, the choice of this par- Hellevik, O. & Bjorklund, T. (1991) Opinion leadership and political
ticular issue may well have impacted on these results in atypical extremism. International Journal of Public Opinion Research, 3, 157–
manner. 181.
A major limitation of this study pertains to the chosen sample. Herr, P.M., Kardes, F.R. & Kim, J. (1991) Effects of word-of-mouth and
It was limited to college students at a single university, and also product attribute information on persuasion: an accessibility diagnos-
gender unbalanced with females representing the majority. Other tic perspective. Journal of Consumer Research, 17, 454–462.
limitations concern the advertisement stimuli used in the survey. Huang, M.-H. (1997) Is negative affect in advertising general or spe-
Future research should include diverse advertisement stimuli with cific? A comparison of three functional forms. Psychology and Mar-
different political messages (both liberal and conservative) in keting, 14, 223–240.
Kim, H., Damhorst, M.L. & Lee, K. (2002) Apparel involvement and
order to more fully understand the effects of political messages on
advertisement processing: a model. Journal of Fashion Marketing and
advertisement and product attitude. Another limitation of this Management, 6, 277–302.
study was the small sample size, which has reduced power to Kim, Y., Park, J. & Kim, Y. (2001) Impact of social advertising claims
detect true relationships among variables. In particular, for a mul- on consumer response: implications for apparel companies. Journal of
tiple regression model with as many independent variables as are Family and Consumer Sciences, 93, 41–45.
incorporated in this study, a larger sample size is required for Laczniak, R., Kempf, D.S. & Muehling, D.D. (1999) Advertising
adequate statistical power. It is also possible that the fact that the message involvement: the role of enduring and situational factors.
advertisements were not created by a professional might have Journal of Current Issues in Advertising, 21, 51–61.
affected the results in some way. Finally, as mentioned earlier, Lord, K.R., Lee, M.S. & Sauer, P.L. (1995) The combined influence
using a diverse selection of fashion products in the advertisement hypothesis: central and peripheral antecedents of attitude toward the
ad. Journal of Advertising, 24, 73–85.
should provide better insight into the relationship between fashion
MacKenzie, S.B., Lutz, R.J. & Belch, G.E. (1986) The role of attitude
opinion leadership and product attitude. toward the ad as a mediator of advertising effectives: a test of
competing explanations. Journal of Marketing Research, 23, 130–
References 143.
Marcus, J. (2003) U.S. students feel the pull of polls after years of
Adkins, L. (1978) How good are advocacy ads? Dun’s Review, June, apathy. The Times Higher Education Supplement, 6 June, 13.
76–77. Mostafa, M.M. (2005) An experimental investigation of the Egyptian
Andersson, S., Hedelin, A., Nilsson, A. & Welander, C. (2004) Violent consumers’ attitudes toward surrealism in advertising. International
advertising in fashion marketing. Journal of Fashion Marketing and Journal of Consumer Studies, 29, 216–231.
Management, 8, 96–112. Nash, N.C. (1995) Benetton touches a nerve and Germans protest. The
Benetton (n.d.) Campaign History. [WWW document]. URL http://press. New York Times, 3 February, D-1.
benettongroup.com/ben_en/about/campaigns/history/ (accessed Purdum, T.S. (2002) The U.S. case against Iraq: counting up the
on 17 May 2006). reasons. The New York Times, 1 October, A19.
Berlyne, D.E. (1960) Conflict, Arousal and Curiosity. McGraw-Hill, Reuters (2000) Sears cites ads in halting Benetton sales. The New York
New York. Times, 18 February, C2.

International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd 97
Fashion advertisements with political content B. Harben and S. Kim

Salmon, C.T., Reid, L.N., Pokrywczynski, J. & Willett, R.W. (1985) The U.S. Census Bureau (2000) Table A-1. Reported voted and registered by
effectiveness of advocacy advertising relative to news coverage. Com- race, Hispanic origin, sex, and age: November 1964 to 1998. [WWW
munication Research, 12, 546–567. document]. URL http://www.census.gov/population/socdemo/voting/
Schlossberg, H. (1992) Members only to include homeless in cause mar- history/htab01.txt (accessed on 17 May 2007).
keting. Marketing News, 26, 6. U.S. Census Bureau (2002) Table 1. Reported voting and registration,
Shah, D.V. & Scheufele, D.A. (2006) Explicating opinion leadership: by sex and single years of age: November 2000. [WWW document].
nonpolitical dispositions, information consumption, and civic partici- URL http://www.census.gov/population/socdemo/voting/p20-542/
pation. Political Communication, 23, 1–22. tab01.txt (accessed on 17 May 2007).
Speckman, K.R. (2004) Reversing youth apathy. The Quill, 92, 30. Vernette, E. (2004) Targeting women’s clothing fashion opinion leaders
Stolle, D., Hooghe, M. & Micheletti, M. (2005) Politics in the super- in media planning: an application for magazines. Journal of Advertis-
market: political consumerism as a form of political participation. ing Research, 44, 90–107.
International Political Science Review, 26, 245–269. Weimann, G. (1994) The Influentials. People Who Influence People.
Summers, J.O. (1970) The identity of women’s clothing fashion opinion State University of New York, Albany, NY.
leaders. Journal of Marketing Research, 7, 178–185.
Taylor, B.J. (2000) Ads for life. Reason, 32, 13.
Tinic, S.A. (1997) United colors and united meanings: Benetton and the
commodification of social issues. Journal of Communication, 47,
3–25.

98 International Journal of Consumer Studies 32 (2008) 88–98 © The Authors. Journal compilation © 2007 Blackwell Publishing Ltd