THE MORAL HIGH GROUND:
PERCEIVED MORAL VIOLATION AND MORAL EMOTIONS
IN CONSUMER BOYCOTTS
by
JOHNNY CHEN
A DISSERTATION
Presented to the Department of Marketing
and the Graduate School of the University of Oregon
in partial fulfillment of the requirements
for the degree of
Doctor of Philosophy
September 20 I0
11
University of Oregon Graduate School
Confirmation of Approval and Acceptance of Dissertation prepared by:
Johnny Chen
Title:
"The Moral High Ground: Perceived Moral Violation and Moral Emotions in Consumer
Boycotts"
This dissertation has been accepted·and approved in partial fulfillment of the requirements for
the Doctor of Philosophy degree in the Department of Marketing by:
Robert Madrigal, Chairperson, Marketing
Lynn Kahle, Member, Marketing
David Boush, Member, Marketing
Robert Mauro, Outside Member, Psychology
and Richard Linton, Vice President for Research and Graduate StudieslDean of the Graduate
School for the University of Oregon.
September 4, 2010
Original approval signatures are on file with the Graduate School and the University of Oregon
Libraries.
© 2010 Johnny Chen
111
IV
An Abstract of the Dissertation of
Johnny Chen for the degree of Doctor of Philosophy
in the Department of Marketing to be taken September 2010
Title: THE MORAL HIGH GROUND: PERCEIVED MORAL VIOLATION AND
MORAL EMOTIONS IN CONSUMER BOYCOTTS
Approved: _
Dr. Robert Madrigal
Prior research has tended to focus on rational (e.g., collective social action and
cost-benefit factors) rather than emotional processes when predicting boycott intention.
The current dissertation proposes that both processes contribute to a boycotting decision
and that each is premised on a perceived moral violation. A model is offered in which
boycott intention is conceptualized as a consumer coping response. Three studies provide
support that moral emotions and cost-benefit factors independently contribute to overall
consumer boycott intention. In Study 1, online survey responses from active boycotters
(N = 121) indicated that participants felt other-condemning moral emotions more acutely
in symbolic boycotts than in non-symbolic boycotts. In Study 2, the theoretical
relationship between perceived moral violation, boycott intention, and boycott behavior
was established in a simple experiment (N = 201). In Study 3, experimental results from
a real world consumer panel (N = 709) indicated that the key to diffusing consumer
boycott intention is counter-message tactics aimed at reducing overall perceived moral
-------------------
v
violation. Path analysis using the data from Study 2 and 3 provided additional insight
into the structure of the proposed model. Other-condemning and self-conscious
emotions, along with perceived boycott benefit (ability to make a difference and self-
enhancement), contributed to boycott intentions whereas cost perceptions played a lesser
role in predicting boycott intention. Comparisons between the hypothesized model and a
set of alternatives supported the proposition that boycott intention may be conceptualized
as a coping behavior. Finally, the results of a path analysis indicated that two individual
difference variables were determinants of perceived moral violation: humanitarian-
egalitarian orientation and negative attitude towards big businesses.
CURRICULUM VITAE
NAME OF AUTHOR: Johnny Chen
GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED:
University of Oregon, Eugene
Oregon State University, Corvallis
DEGREES AWARDED:
Doctor of Philosophy, Marketing, 2010, University of Oregon
Master of Business Administration, Marketing, 2000, University of Oregon
Bachelor of Science, Computer Science, 1994, Oregon State University
AREAS OF SPECIAL INTEREST:
Experiential Consumption (Sports and Entertainment)
Prosocial Consumer Behavior
Withholding Consumption
Structural Equation Modeling
PROFESSIONAL EXPERIENCE:
Graduate Teaching and Research Fellow, Department of Marketing, Lundquist
College of Business, University of Oregon, Eugene, 2005-2010
Product Line Marketing Engineer, Sf., Intel Corporation, Hillsboro, Oregon,
2000-2005
Network Specialist, Corporate IT, Intel Corporation, Hillsboro, Oregon, 1994-
2000
VI
Vll
GRANTS, AWARDS AND HONORS:
AMA Sheth Foundation Doctoral Consortium Fellow, Columbia, MO, 2008
Merle King Smith Marketing Scholars Award, 2005-2010
Calvin Reed Smith Research Grant, 2005
PUBLICATIONS:
Gurel-Atay, E., Xie, G., Chen, J., & Kahle, 1. R. (2010). Changes in social values
in the United States, 1976-2007: "Self-respect" is on the upswing as "sense of
belonging" becomes less important. Journal ojAdvertising Research, 50(1),
57-67.
Madrigal, R., Chen, J., LaBarge, M., & Sagara, N. (2009). Consumers' Response
to Advocacy Advertising: A Process Model of Consumer Skepticism,
Empathic Response, and Prosocial Behavior. Advances in Consumer
Research, 36,731-732.
Madrigal, R., & Chen, J. (2008). Moderating and mediating effects of team
identification in regard to causal attributions and summary judgments
following a game outcome. Journal ojSport Management, 22(6), 717-733.
Chen, J., & Madrigal, R. (2008). A Bibliographic Survey of Experiential
Consumption Research. Advances in Consumer Research, 35,976-977.
Vlll
ACKNOWLEDGEMENTS
I would like to thank my committee members-Bob Madrigal, Dave Boush, Lynn
Kahle, and Robert Mauro-for sharing with me their time, experience, and insight to help
steer my dissertation in the right direction. Their kindness and encouragement made this
dissertation a fantastic experience. lowe a special debt of gratitude to my advisor, chair,
and mentor, Bob Madrigal, for shepherding me through the doctoral process from day
one. He has been a role model for me since I first met him in the MBA sports marketing
program.
I would like to thank my fellow doctoral cohorts Ian Parkman, Courtney
Boerstler, and Namika Sagara for a phenomenal shared doctoral experience. I would also
like to thank my fellow co-authors Monica LaBarge, Eda Gurel Atay, and Guang-Xin Xie
for the pleasure of collaborating on some very impactful research. These fellow travelers
help defined a period of my life full of happiness and purpose.
IX
TABLE OF CONTENTS
Chapter Page
I. INTRODUCTION 1
II. LITERATURE REVIEW 8
Consumer Boycotts 8
Consumer Boycott Models 11
Emotions and Boycott Intention 17
Hypothesized Model 24
Alternative Models 26
Counter-Messaging Tactics 28
III. METHODOLOGY 31
Overview 31
Key Model Variables 32
Individual Difference Variables 36
Model Testing and Alternative Models 38
Study 1: Exploratory Online Study of Active Boycotters 41
Study 2: Conceptual Model Testing 43
Study 3: The Effect of Counter-Messaging on Boycott Intention 46
IV. RESULTS 48
Study 1 48
Study 2 60
Study 3 76
V. DISCUSSION 97
Perceived wloral Violation 100
Moral Emotions 103
An Affect-Driven Model. 104
Limitations and Future Research 107
Postscript 110
Chapter
x
Page
APPENDICES 111
A. SUMMARY OF PRIOR BOYCOTT RESEARCH 111
B. STUDY 1 SURVEY INSTRUMENT 113
C. STUDY 2 SURVEY INSTRUMENT & STIMULI 123
D. STUDY 3 SURVEY INSTRUMENT, PRO-BOYCOTT MESSAGE,
AND COUNTER-MESSAGES 139
E. MEDIATION WITH CONTINUOUS MEDIATOR AND
DICHOTOMOUS OUTCOME (SPSS CODE) 158
F. STUDY 2 ML AND BAYESIAN PARAMETER ESTIMATES 161
G. STUDY 3 ML PARAMETER ESTIMATES 164
BIBLIOGRAPHy 167
Xl
LIST OF FIGURES
Figure Page
1. A two-process consumer boycott modeL 25
2. Hypothesized model. 27
3. Affect-driven model. 27
4. Reasoning-driven model. 28
5. Alternative models 39
6. Source of boycott recruitment. 51
7. Likelihood of ending boycott if target concedes to demands 52
8. Boycott intention mediates perceived moral violation 68
9. Path diagram for hypothesized process model and estimated
path coefficients 70
10. Path diagram for individual process components (Study 2) 75
11. Profile of manipulation check indices for counter-message tactics 86
12. Effect of counter-messages on boycott intention (BI) and perceived
moral violation (PMV) 88
13. Path analysis of hypothesized process model using Study 3 data 89
14. Structural configuration of alternative models (Study 3) 92
15. Path analysis of individual emotional and cost-benefit factors (Study 3) 95
16. Best fitting alternative model (Model 5) 105
----------------- ------_._----
XlI
LIST OF TABLES
Table Page
1. Perceived Moral Violation Measure 32
2. Boycott Intention Measure 33
3. Moral Emotions Measure 33
4. Cost-Benefit Factors 35
5. Humanitarian-Egalitarian Scale 37
6. Product Category Stimuli Presented to Participants 44
7. Sample Characteristics (Study 1-3) 49
8. Individual Differences (Study 1) 50
9. Boycott Targets and Motivations 54
10. Differences in Key Measures for Instrumental vs. Symbolic Boycotts 57
11. Descriptive Statistics and Correlation Matrix (Study 2) 63
12. Means Comparisons for Non-Boycotters and Boycotters (BOyCOTT) 64
13. Logistic Regression for Perceived Moral Violation and Boycott Intent as
Predictors of Boycott Behavior 66
14. Fit Statistics for Boycott Behavior Models 66
15. Observed and Predicted Frequencies of Boycott Behavior (BOyCOTT) 67
16. Decomposition ofIndirect Effects from Path Analysis (Study 2) 72
17. Aiternative Models Fit Statistics (Study 2) 73
18. Descriptive Statistics (Study 3) 80
19. Experimental Conditions and Counter-Message Tactics 82
Table
X111
Page
20. Summary of Manipulation Check Indices, Items, and Corresponding Tactics 84
21. Means for Counter-Message Manipulation Indices (CMl-4) 85
22. Summary of Planned Contrasts for Boycott Intention (BI) 87
23. Decomposition ofIndirect Effects from Path Analysis (Study 3) 90
24. Alternative Models Fit Statistics (Study 3) 93
CHAPTER I
INTRODUCTION
Emotions and Consumer Boycotts
Consumer boycotts represent a threat to consumer brands. Activist groups can
organize and execute effective boycotts with remarkable ease against large multinational
brands using the Internet and mass media (Sen, Gurhan-Canli, & Morwitz, 2001; N. C.
Smith, 2005). John and Klein (2003) proposed that boycotts in the United States have
increased about fourfold from the mid-1980s to the mid-1990s. Klein et al. (2004)
estimated that 42% ofleading companies and 52% of major brands were facing consumer
boycotts of one form or another. Companies worry about boycotts because one third of
boycotted firms have conceded to boycott demands (Klein, et aI., 2004). A boycott
threatens not only a firm's sales and stock prices (Miller & Sturdivant, 1977; Pruitt &
Friedman, 1986; Pruitt, Wei, & White, 1988), it also damages brand equity and requires
firms to divert managerial and financial resources for public relations and damage control
(Kozinets & Handelman, 1998). Worse, the damage to the brand lingers beyond the direct
effects of the boycott because both boycotters and non-boycotters are likely to devalue
their perception of the brand (Klein, et aI., 2004).
Boycotts are collective social actions that use the marketplace to apply social,
economic, and political pressures against an offending party. Consumers participate in
boycotts against a person, a firm, an industry, or even a nation to remedy a personal or
2social injustice. Historically, boycotts have been powerful tools for unions and political
activists. Famous examples of historical boycotts include the American tea boycott of
1773 (which led to the Boston "Tea Party"), Gandhi's British textiles boycott, and the
Montgomery bus boycott. However, the Internet and mass media have dramatically
changed the nature of modem consumer boycotts (Friedman, 1999; N. C. Smith, 1990).
Unlike historical boycotts, modem boycotts are increasingly symbolic, prosocial,
and emotional. In addition, they rely on voluntary participation by initially disinterested
consumers (Friedman, 1999). Symbolic boycotts highlight social injustices and utilize
moral pressure in addition to economic or political pressure to force the offending party
to take corrective action. The most common type of symbolic boycott is a media-oriented
boycott that supports a prosocial or altruistic cause. The goal of these boycotts is to help
underrepresented victims or the environment. Symbolic boycotts are very different from
union-sponsored boycotts, which primarily benefit those who initiated and participated in
the boycott. By contrast, symbolic boycotts are designed to benefit others. Some
examples of symbolic boycotts include the Stop Killer Coke campaign, Global
Exchange's slave chocolate boycott, and PETA's fur boycott. The act of withholding
consumption in this context is a conspicuous prosocial action similar to charitable giving
or volunteering. Consumers become emotionally involved in the welfare of others
through boycotting. Thus, boycotting is often a deeply emotional and personal
experience.
Three perspectives on individual consumer decisions to participate in boycotts
exist in the marketing literature: purchase votes (Dickinson & Hollander, 1991),
3collective social actions (John & Klein, 2003; Klein, et ai., 2004; Sen, et ai., 2001), and
individual emotive (Kozinets & Handelman, 1998). The purchase vote perspective
suggests that consumers aggregate their "purchase" votes to reward firms that maximize
societal interests. This form of collective action is consistent with ethical consumption
(Cooper-Martin & Holbrook, 1993). The vote analogy is predicated on a self-correcting
free market system where firms that do not maximize societal interests will be punished.
This view addresses the ethical dimension of boycott participation but is fundamentally
economic in nature.
Collective action suggests that consumers make an economic assessment of
personal cost and benefit to decide whether to participate in a boycott. So long as benefits
outweigh costs, it is beneficial to act collectively toward boycott goals. This stream of
research views boycotting as a form of prosocial behavior. The final perspective
considers boycotting as an individual emotive expression. The only research that has
directly addressed the fundamental issue of moral violation and the expression of
emotions is a netnography by Kozinets and Handelman (1998). Two emergent themes
challenged the previous views that boycotting is a collective social action. Kozinets and
Handelman found boycotting was a "morally transforming behavior" and "a personal,
rather than communal, act" (p. 477). Little empirical research has followed up on the
emotional view of individual motivation to participate in consumer boycotts. This is
surprising considering the most successful boycotts tend to be "cognitively simple and
emotionally appealing" (Friedman, 1999, p. 198).
4The different boycott models need not be mutually exclusive. Like other
consumer decisions, it is likely that a combination of moral, emotional, and economic
factors motivate consumers to participate in boycotts. The collective social economic
models and the emotional and morally transformative models may be integrated because
a moral violation or transgression is usually at the heart of a consumer boycott. Sen et al.
(2001) and Klein et. al (2004) empirically established that moral transgressions as a result
of a company's egregious actions are antecedent to cost-benefit evaluations and boycott
intention. Klein et al. (2004) defined this transgression as "the belief that a firm has
engaged in conduct that is strikingly wrong and that has negative and possibly harmful
consequences for various parties (e.g., workers, consumers, society at large)" (p. 96).
This definition coincides with the definition of perceived moral violation, a cognitive
appraisal also known as legitimacy, norm/self concept comparability, and normative and
moral comparability (Jolmson & Stewart, 2004; Lazarus, 2001; Mauro, Sato, & Tucker,
1992; Roseman & Smith, 2001; Scherer, 1997; C. A Smith & Ellsworth, 1985; C. A
Smith & Lazarus, 1993). Perceived moral violations elicit moral emotions, which in turn
lead to the formation of coping intentions and responses (Haidt, 2003). Boycott behavior
may be conceptualized as one such consumer coping behavior (Duhachek, 2005) because
it is motivated by consumers' appraisal of perceived moral violation and emotional
reactions in response to an act perpetrated by a company or brand. Thus, a consumer's
perception of moral violation is related to two complementary processes that may result
in the same behavior.
5Research Objectives
The current research attempts to advance three main ideas in consumer boycott
research. First, the increased popularity of symbolic boycotts requires a reevaluation of
current definitions of boycott participation because the traditional economic definition of
withholding purchase does not fully represent the range of actions used by boycott
participants. Consumers are likely to participate in boycotts through a mix of actions that
includes private discussion of dissatisfaction, consideration and intentions to boycott,
actual withholding purchase and consumption, brand switching, public discussion of
dissatisfaction, and activism (boycott promotion and recruitment). Questions that identify
specific boycott actions should replace ambiguous statements such as "how likely are you
to boycott" that are often used in research. The current research conceptualizes a broader
view of consumer boycott intention that is applicable to a greater variety of boycotts.
Second, John and Klein (2003) framed non-rational factors in boycott (for
example, emotions) in purely economic terms, but there may be insights gained by
examining emotional and cost-benefit processes separately. Emotional expression and
rational cost-benefit factors are complementary mechanisms in a wide variety of
psychological processes, and each is likely to contribute to a consumer's willingness to
participate in a boycott. Thus, a two-process model that encompasses emotional and cost-
benefit factors may be a better model for describing consumer boycott participation.
Third, boycott intention is conceptualized here to be a form of emotional coping
that is triggered by a moral evaluation of an unfair or unjust situation that affects
someone else. Prior literature has linked the moral emotions of outrage and guilt to
6boycotts (see Appendix A). Applying Haidt's (2001,2003) moral emotion framework to
boycott situations may identify additional moral emotions that are relevant to boycotting.
Further, the investigation of moral emotions provides new avenues for consumer research
in the domains of ethical consumption and corporate social responsibility.
Three studies are presented to test a two-process model for consumer boycott
participation. First, an exploratory survey was used to assess how perceived moral
violations, emotions, and cost-benefit factors are represented in real world boycotts.
Second, participants are asked to attend to their emotions or cost-benefit evaluation of
several boycott scenarios in an experiment in order to examine the structure of a proposed
conceptual model. In the final study, the marketing application of the conceptual model
was assessed. A consumer panel was employed to test the effectiveness of a variety of
counter-messages on reducing boycott intention.
Organization of Dissertation
Chapter II describes existing consumer boycott models and discusses research
related to the appraisal theory of emotion and the moral emotion framework. The chapter
concludes by proposing a two-process model that extends current economic models by
incorporating a complementary emotional process.
Chapter III describes the research design and execution of three studies, and
provides a summary of key independent and dependent variables. Study 1 provided a
descriptive profile of real world boycott participants based on key dependent variables
that will be used in later studies. Study 2 examined the relationship between perceived
moral violations, boycott intention, and boycott behavior in an experiment. Study 3
examined the efficacy of counter-messages proposed by Klein et al. (2004) on real
consumers. In addition, the experimental manipulations in Study 2 and Study 3 provided
the needed variance in the data to model and test the proposed two-process model and
alternative models using path analysis.
In Chapter IV, the results of each study as well as path analyses are presented.
Chapter V concludes with a summary of key findings, theoretical and practical
implications, limitations, and potential future research.
7
8CHAPTER II
LITERATURE REVIEW
Consumer Boycotts
A consumer boycott is the purposeful withholding of consumption of a specific
product, a category of products, or a brand to exert influence on an offending firm
(Friedman, 1999; Sen et aI., 2001; N. C. Smith, 1990). Boycotts are thought to be
triggered when a consumer perceives a company's or industry's actions or practices as
violating moral norms. In retaliation, consumers apply economic and social pressure on
the offending business en masse. Collective action increases buyer power directed against
a firm, an industry, or even a nation (John & Klein, 2003). Boycott success is declared if
the boycott leads to a cessation or remedy of the egregious act or practice (N. C. Smith,
1990).
Boycotts can be either instrumental or symbolic. A marketplace-oriented boycott is
primarily instrumental because it applies socioeconomic-political pressures and is
typically instigated by members of the affected party. By contrast, symbolic boycotts
differ from instrumental boycotts in three ways. First, a symbolic boycott is a moral
expression of consumer disapproval and disassociation (N. C. Smith, 1990). Symbolic
boycotts seek compliance through moral as well as socioeconomic-political pressure.
They have historically been used as a weapon by the weak and are designed to promote
9social justice, and to protect the rights of the powerless and disenfranchised segments of
society (N. C. Smith, 1990). Second, symbolic boycotts have been hypothesized to reflect
individual expressions of moral outrage (John & Klein, 2003). Symbolic boycotts
principally serve to provide an avenue for the protesting organization and boycott
participants to vent frustration, often with broad unspecified goals (Friedman, 1999).
Finally, the primary emphasis of a symbolic boycott is to damage the target brand, its
image, or corporate morale. Modem symbolic boycotts call attention to specific corporate
practices instead of promoting specific sociopolitical goals.
However, as labor and consumer price boycotts recede into the past, boycotts have
more recently focused less on economic necessities and more on prosocial or ethical
concerns. The shift from instrumental to symbolic and from marketplace to media has
had a profound effect on consumers' brand perceptions and their beliefs about what
boycotts can accomplish (Friedman, 1999). Yet, in spite of this shift, most consumer
research has focused on instrumental boycotts that seek sociopolitical or price changes
(Friedman, 1991, 1999). Friedman (1991) suggested the need to tum attention to studying
symbolic boycotts, as they have grown in prominence and popularity.
Symbolic boycotts are most often implemented through the media. Rather than
drive short-term economic impact in the marketplace, the media is enlisted to damage
brands, shift consumer sentiments, and distract corporate attention. However, it is
possible for a symbolic boycott to affect an offending company's bottom line as a result
of the long-term damage to the company's brand (N. C. Smith, 1990). Thus, activists
favor using media-oriented boycotts against highly visible brands. Market leaders are
10
more vulnerable because they have more brand eq~ity at stake (Brown & Dacin, 1997;
Sen & Bhattacharya, 2001).
Most activist groups calling for consumer boycotts commit few, if any, resources
to their execution in the marketplace. Instead, activist groups are increasingly adroit in
implementing pro-boycott messaging through cost-effective media such as the Internet,
press releases, and staged events (John & Klein, 2003). Media messages are designed to
generate strong emotions that will motivate consumers to join the boycott. In most
executions, people are asked to make a personal sacrifice by withholding consumption of
a preferred brand as the morally correct thing to do. In response, modem companies are
equally adept at managing public relations through counter-messaging. For example,
Nike's public relations team grew by 70 members after Workers Rights Coalitions
allegation of worker abuses and poor working conditions in its plants abroad and
threatened to boycott (Beder, 2002). Not only did the company issue public release
statements, but they also initiated a public relations campaign of its own to improve the
perception of the company's actions in consumers' eyes.
Boycott actions represent a spectrum of related behaviors (Friedman, 1999; N. C.
Smith, 1990). In previous consumer research, a boycott is often associated only with
withholding purchase or is ambiguously defined (e.g., Klein et aI., 2004; Sen et aI.,
2001). In reality, protest, complaint, word of mouth, and activism are actions equally
crucial to boycott success. Differentiating between these actions is important in boycott
research because each action requires a different level of commitment from consumers.
Six distinct consumer boycott actions with escalating militancy are as follows:
11
1. Consumer discusses dissatisfaction in private.
2. Consumer considers and has intent to withhold patronage.
3. Consumer actually withholds patronage.
4. Consumer seeks out and buys competitors' products.
5. Consumer publicizes dissatisfaction with the target.
6. Consumer publicizes the boycott itself and persuades others to join.
Boycott behavior is different from exit and brand switching (A. O. Hirschman, 1970).
Consumers are buying from a competing brand to exert economic pressure and social
censure. The decision to switch is deliberate and purposeful. Often, the consumer would
resume patronage of the offending brand if the initial offense were remedied. A positive,
or whitelist, boycott (also known as a buyeatt) is based on directing consumers to
purchase a brand other than the target brand, therefore indirectly boycotting the offending
brand.
Rather than viewing boycotting as a monolithic construct, modern boycotts are
more complex because their objectives are symbolic rather than economic. Thus, the
measurement and description of boycott behaviors need to be more than simply
withholding purchase in order to fully capture the scope and range of modern
boycott behaviors.
Consumer Boycott Models
Three different views have been presented in research on consumers' decision to
participate in boycotts. First, consumers use purchase votes to favor firms with a
12
preferred societal impact (i.e., Dickinson & Hollander, 1991; Klein et aI., 2004; Sen et
aI., 2001). Second, boycotts are collective social actions based on consumers' cost-
benefit evaluations (i.e., Innes, 2006; John & Klein, 2003; Klein, et aI., 2004; Sen, et aI.,
2001). Third, boycotts are individual emotive expressions (Kozinet & Handelman, 1998).
All three views support the notion that morality plays a central role in a person's desire to
participate in a boycott. The collective action and individual expression views also imply
that there is an affective link between the perceived moral violation arising from the
egregious act by a company and subsequent boycott intention. Yet, to date, little research
has examined the nature of this affective link.
Purchase Votes
The purchase vote perspective suggests that consumers aggregate their
"purchase" behavior to reward firms that maximize societal interests (Dickinson &
Hollander, 1991; N. C. Smith, 1990). This view hints at the morality of boycott
participation but uses an economic model to do so. The vote analogy is predicated on a
self-correcting free market system where firms failing to maximize societal interests will
be punished. This definition is problematic because it requires a vote be cast for one firm
over another; yet, a third option to cast no vote for either firm is also available. When
consumers simply abstain, it is unclear how this action will lead to the cessation of a
specific firm's egregious behavior. Companies cannot be certain if a drop in sales is in
response to an egregious act or other factors. Without specific ties between sales and
unethical action, a company simply would not be aware that it has committed an
13
egregious act, or may interpret its actions as being ethical in the absence of specific
protest or consumer response.
Although the vote analogy is more vivid, it does not adequately capture the
essence of withholding purchase and consumption in the boycott context. The purchase
votes analogy is more consistent with an active ethical consumption experience. Cooper-
Martin and Holbrook (1993) differentiated between active and passive ethical
consumption experiences. Purchase votes are akin to active ethical consumption
experiences requiring the consumer's overt participation. Withholding consumption is a
passive consumption experience that is a reactive response of pursuing good by not
buying harmful items (Cooper-Martin & Holbrook, 1993). Purchase voting is more
consistent with buycotts than with boycotts, and should potentially be treated as a
separate set of behaviors.
Cost-Benefit Models
The collective action research views boycott participation as a consequence of
deliberative and rational assessments of personal costs and benefits. So long as benefits
outweigh costs, it is beneficial for consumers to act collectively toward boycott goals.
Sen et al. (2001) conceptualized boycott participation using social dilemma theory. In a
social dilemma, the interests of the individual are at odds with the interests of the group.
Consumers must choose between maximizing either selfish or collective interests. In this
model, three factors determine a consumer's decision to participate in a boycott: (a) the
perceived likelihood that a boycott will be successful; (b) the perceived compliance and
14
participation of others; and (c) the associated boycott cost to the individual consumer
weighed against the benefit to a group. Sen et al. (2001) concluded that people's
uncertainty about boycott success was a key barrier to participation. Specifically,
perceived success likelihood was found to mediate the direct effects of perceived
efficacy, expectation of overall participation, and pro-boycott message frame on boycott
likelihood.
Klein et al. (2004) extended Sen et al. 's (2001) model by conceptualizing social
boycotts as a helping behavior. The authors proposed four cost-benefit factors that predict
boycott participation: (a) making a difference, (b) self-enhancement, (c)
counterarguments, and (d) constrained consumption. These factors incorporate Sen et
al.' s factors. For example, Klein et al. 's factor for making a difference is consistent with
Sen et al.'s perceived likelihood of success. The self-enhancement factor included
elements of social normative influence (i.e., pressure of family/friends, uncomfortable if
seen by others). Counterarguments and constrained consumption are consistent with
social dilemma tensions outlined in Sen et al.' s model. Klein et al.' s unique contribution
to the cost-benefit model was the inclusion of the self-enhancement factor, which
combined the need to avoid guilt and the need to feel better. Klein et al. found that self-
enhancement was positively related to perceived egregiousness and individual boycott
participation. Participants were more likely to boycott if they believed they could make a
difference and if they felt better about themselves by taking action. In addition,
participants were less likely to boycott if strong counterarguments were made against
------- --- --------
15
participation and if the constrained consumption costs were greater than the perceived
benefit of boycotting.
Klein et al.'s (2004) Awareness-Egregiousness-Boycott (AEB) model was notable
because the authors called attention to three important ideas. First, perceived
egregiousness was the starting point for boycott consideration. Klein et aI. defined
egregiousness as a "belief that the firm has engaged in conduct that is strikingly wrong
and has negative and possibly harmful consequences for various parties (e.g., works,
consumers, society at large)" (p. 96). This definition is consistent with a moral or ethical
violation (Haidt, 2003).
Second, Klein et aI. noted that "consumers encounter an initial trigger event that
engenders negative arousal" (p. 93). This suggests that the deliberative cost-benefit
analysis associated with boycotting is the result ofan affective trigger. Sen et aI. (2001)
also suggested that future research needs to examine if positive and negative affect
moderates the decision to participate in boycotts.
Finally, Montada and Schneider (1989) found that prosocial action was often an
emotional response to a social injustice. Despite framing boycotts as a form of prosocial
behavior, emotions associated with prosocial behavior in prior research, such as anger
(e.g., Bagozzi & Moore, 1994), guilt (e.g., Basil et aI., 2006), empathy, and sympathy
(e.g., Batson, 1998; Eisenberg, et aI., 1989) have not been prominently featured in
research on boycotting.
16
Individual Expression
The personal commitment to social justice and prosocial causes associated with
symbolic boycotts suggests that emotions may play an important role in many boycott
decisions. Kozinets and Handelman's (1988) online exploratory study suggested that the
act of boycotting was a "complex emotional expression of individuality" (p. 475). The
ethnography yielded two emergent themes that challenge the collective social action and
purchase vote perspectives of boycott participation. First, boycotters consider their
participation to be a personal rather than communal act. Boycott participation was valued
for its ability to convey the boycotter's uniqueness and lack of conformity. Thus, the
expression of an extended-self was achieved not only by what was purchased and
consumed, but also by what was not consumed (Belk, 1988).
The second emergent theme viewed boycott participation as a morally
transformative behavior. Boycott participation allows the participant to be differentiated
from the mainstream and to stand above them morally. Both themes suggest that strong
individual emotional responses are antecedent to a consumer's decision to withhold
consumption. To date, no research has tested for this possibility.
Klein et al.' s (2004) AEB model provides the basic framework for consumer
boycott behavior, but it does not fully acknowledge the crucial role that emotional
expression plays in motivating consumer boycotts. Beyond economic reasons, consumers
also express themselves emotionally in the marketplace (O'Shaughnessy &
O'Shaughnessy, 2003). Thus, it is possible that a consumer boycott model that includes
ethics, cost-benefit evaluations, and emotional expressions as key contributors to boycott
17
intention may be more generalizable across a range of boycott situations than models that
simply include economic factors.
Emotions and Boycott Intention
Emotions are responses to perceived changes, threats, or opportunities in the
environment that guide behavior (Ekman, 1999; Frijda, 1988). In consumer research,
emotions are thought to influence information processing, motivate or inhibit action, and
contribute to experience or satisfaction (e.g., Bagozzi, Gopinath, & Nyer, 1999; E. C.
Hirschman & Holbrook, 1982). Although considerable research has accumulated since
Holbrook and Hirschman's (1982) call to examine the central role emotion plays in
consumption, recent emphasis on cognitive appraisal theory is looking at the role emotion
plays in consumer decisions in a more detailed and process-driven way. Emotions and the
actions they elicit are linked as consumer coping responses (Duhachek, 2005; Duhachek
& Iacobucci, 2005; Duhachek & Oakley, 2007). More importantly, the flexibility of the
cognitive appraisal theory as a general theory of emotions provides a potential framework
to link perceptions of moral violations, emotions, and boycott intention.
The process of cognitive appraisal encompasses three stages (Johnson & Stewart,
2004; Lazarus, 1991): (a) situational awareness, (b) appraisals and emotions, and (c)
coping process. When confronted with a situation or event (i.e., boycott), a consumer
evaluates (appraises) the event subjectively based on personal knowledge, goals, and
idiosyncrasies. The consequences of these appraisals are discrete emotions (Lazarus,
2001 ; Roseman & Smith, 2001). Finally, behavioral tendencies associated with specific
18
emotions help a person to adapt to the situation or event. These tendencies may result in
overt behaviors, given sufficient opportunity and capability. This last stage is known as
the coping process. Boycott behaviors are proposed to be a mechanism for coping with
emotions involved in the boycott situation. A description of each stage follows.
Situational Awareness
The antecedent to appraisal may be described as situational awareness, and may
be thought of as the interaction between the environment or situation and individual
knowledge, goals, and idiosyncrasies (Johnson & Stewart, 2004). This may explain why
different people feel different emotions given the same situation. Two moderators
thought to affect a consumer's situation awareness in the case of boycott decision making
are humanitarian-egalitarian orientation and negative attitudes toward big companies.
In social justice research, Katz and Hass (1988) described "adherence to the
democratic ideals of equality, social justice, and concern for the others' well-being" (p.
894) as the humanitarianism-egalitarianism (HE) orientation. This orientation varies from
person to person and inherently colors a person's perception of what is fair and just. In
studies related to racial equality issues, Katz and Hass suggested that an HE outlook
created the commitment to justice and sympathy for the underdog. When confronted with
a company's egregious act that harms those unable to protect themselves, people who
hold a strong HE outlook should be more sensitive to the moral violation.
The corporate social responsibility literature suggests another moderator likely to
affect situational awareness is a consumer's attitude toward big business (Webster, 1975).
19
Because boycotts typically target large multinational corporations, participants with
negative attitudes toward big business may be more likely to perceive moral
transgressions in the actions of big business and be more suspicious of big business
communications in boycott situations. Therefore, people with negative perceptions of big
business may be more likely to take actions against them.
Appraisal and Discrete Emotions
Cognitive appraisals are direct antecedents to discrete emotions (Madrigal, 2008;
Roseman & Smith, 2001; Scherer, 1997,2001). Emotions are differentiated by patterns of
appraisals along a common set of dimensions. Although past research does not
conclusively agree on the specific number and labels of appraisal dimensions, five
dimensions appear consistent in the literature (Johnson & Stewart, 2004; MacInnis & de
Mello, 2005; Mauro, et aI., 1992; C. A. Smith & Ellsworth, 1985): (a) relevance, (b) goal
congruence, (c) moral or normative comparability, (d) agency, and (e) future expectancy.
Relevance is the degree of importance a particular situation has for a person. The
greater the relevance, the more likely a situation will elicit an emotional response. Goal
congruence is the evaluation of a positive or negative outcome for a situation. Goal
congruence typically determines the valence and intensity of an emotion (positive or
negative). Moral comparability incorporates evaluations of fairness, justice, legitimacy,
and social standards. The appraisal of moral comparability is especially important in
differentiating moral emotions where the welfare of others is at stake (Johnson &
Stewart, 2004). Agency incorporates the concepts of accountability and responsibility.
20
Specifically, blame or credit is given to the perceived cause of a given event or situation.
Future expectancy is an assessment of how things will turn out, including effective and
ineffective coping (Lazarus, 1991). In addition, future expectancy is related to certainty
(Johnson & Stewart, 2005; MacInnis & de Mello, 2005) and controllability (Lazarus,
1991).
The appraisal of moral comparability deserves special attention in the boycott
context. At the heart of a consumer boycott is the perception of a moral violation or
transgression committed by the company. This is especially true for symbolic boycotts
that seek compliance through moral pressure. Without a strong sense of moral violation
or transgression, insufficient moral anger or guilt exists to provide the "heat" needed to
trigger an intention to boycott. The individual moral emotions require an additional
explanation because they provide clues to how some emotions may be expressed through
certain boycott behaviors.
Moral emotions. Moral emotions are a subset of general emotions that are linked to
the prosocial interest or welfare of society as a whole, a group, or a third party (Haidt,
2003). Often, moral emotions provide the "motivational force" to do good and avoid
doing bad (Tangney, Stuewig, & Mashek, 2007). In boycotts that seek to redress social
ills or to help others, moral emotions are a guiding mechanism in altruistic and helping
behaviors. It follows then that moral emotions will be antecedent to an intention to
participate in a boycott.
Moral emotions can be differentiated into two groups based on the appraisal of
agency: other-condemning and self-conscious. The other-condemning family includes
21
anger, contempt, and disgust. The self-conscious group includes guilt, shame, and
embarrassment. Of these, anger and guilt represent the most prototypical moral emotions
leading to greater prosocial action tendency in the support of another (Haidt, 2003).
Anger is a function of other-blame with the attributes of being motivationally relevant but
incongruent and predicated on the accountability of a third party (C. A. Smith, Haynes,
Pope, & Lazarus, 1993). Guilt is a function of self-blame with the attributes of being
motivationally relevant but incongruent and predicated on self-accountability (C. A.
Smith, et aI., 1993). A description of each follows.
Anger in this context may be better described as moral anger or moral outrage as a
result of a violation of fairness and justice. Anger motivates direct retaliation against the
party responsible for the perceived egregious act, even if a person has no ties to the
victimized group (Haidt, 2003). The need for revenge, humiliation, attack, and
punishment are examples ofretaliatory tendencies arising from righteous anger.
Oppression, exploitation, and racism are well-known examples of situations where
righteous anger is expressed. Thus, anger should be a predictor of the intention to boycott
(Friedman, 1999; Klein et aI., 2004; N. C. Smith, 1990).
Guilt in the context of boycotting refers to an awareness of one's own actions that
violate moral rules that cause harm and suffering to others (Haidt, 2003; Hoffman, 1987).
As such, it may be better described as moral guilt. Moral guilt is the only self-conscious
moral emotion that motivates helping behavior (Haidt, 2003). Tangney et aI. (2007)
suggested that moral guilt facilitates empathic processes and encourages the use of
constructive strategies for coping with anger and reducing aggressive behaviors. Guilt-
22
related action tendencies favor avoidance, withdrawal, or reparation rather than attack or
punishment.
An emotion closely related to guilt is regret. Whereas guilt refers to harming
another, regret is self-blame for harming oneself. It is possible that guilt and regret are
confounded in boycott situations where consumption of a product by a company that has
committed an egregious act against others can be perceived as not only harmful for others
but also harmful to one's self-esteem or beliefs, especially in the context of modern
consumers who believe their consumption is sustainable or ethical. Because both guilt
and regret are negative emotions with very similar coping outcomes (Passyn & Sujan,
2006), it may be useful to examine guilt and regret together in the boycott situation.
Coping Responses
Coping is the final stage of the appraisal process. Boycott behaviors are considered
a form of coping in the current research. According to Lazarus (1991), "Coping is the
psychological analogue ofaction tendencies" (p. 830). Coping occurs in response to
emotional stress because of situational or environmental change. Consumers employ
coping strategies to deal with emotionally stressful consumption situations. Duhachek
(2005,2007) and Duhachek & Iacobucci's (2005) meta-analysis of the coping literature
resulted in three categories of consumer coping strategies: (a) active coping, (b)
expressive support-seeking, and (c) avoidance.
When consumers utilize active coping, they take action both cognitively and
behaviorally, including "engaging in rational thinking" (Duhachek, 2005). Boycott
23
behaviors consistent with active coping include seeking out substitutes, speaking out,
activism (consumers publicize the boycott and recruit peers to join the boycott), and
public displays of dissatisfaction. In this case, withholding purchase is an active coping
mechanism where the intention is to punish. Expressive support-seeking engages social
resources and emotional venting. This includes both public and private discussions.
Avoidance coping seeks ways to avoid the stressor. Moral guilt may be expressed as a
form of avoidance coping. Boycott participants are likely to discuss privately, reflect on
the issues, and withhold consumption to avoid feeling guilt or as a way to begin making
reparation.
In summary, the cognitive appraisal process suggests that consumer
interpretation ofa boycott situation elicits an emotional response. These emotions can
engender a variety of behavioral intentions. The stronger the emotions, the more likely
will be the consumer's need to express or cope with the emotion. Thus, emotions provide
a viable process linking perceived moral violations to subsequent boycott intention.
Emotions and emotional processes are natural extensions of cost-benefit models
like the Klein et al.'s (2004) AEB model. In prior research, moral outrage (Friedman,
1985, 1991, 1996, 1999; N. C. Smith, 1990) and guilt (Klein et aI., 2004; Kozinets &
Handelman, 1998; N.C. Smith, 2005) provided an unspecific and vague precondition to
boycotting, while cost-benefit factors were identified as direct contributors to boycott
intention. However, little research has investigated the direct contribution of emotions to
boycott intention. In addition, emotions are important because they provide another way
to reconcile moral judgments into the overall boycott model.
24
Hypothesized Model
Sen et al.'s (2001) and Klein et al.'s (2004) collective action models provide the
foundation for the proposed two-process model tested in the current dissertation.
Specifically, the perceived moral violation arising from a company's actions was
positively related to consumer boycott intention by Klein et al. (2004), whereas Sen et al.
(2001) demonstrated that boycott intention was positively related to boycott behaviors.
The two-process model proposed here incorporates these effects, but also includes the
intervening variables of moral emotions and cost-benefit processes between perceived
moral violation and boycott intention. Consequently, the hypothetical model shown in
Figure 1 proposes that intention to boycott is the result of both an emotional and a
rational process. Specifically, an appraisal of moral violation elicits both other-
condemning and self-conscious moral emotions (Haidt, 2003), and both types of moral
emotions are predicted to increase intention to boycott.
When a consumer perceives that a company's actions have violated a moral norm,
the consumer is expected to make a reasoned decision to dissociate from the company
after a cost-benefit analysis (John & Klein, 2003; Klein, et aI., 2004; Sen, et aI., 2001).
Perceived moral violation is proposed to be positively related to benefit and negatively
related to cost. It is likely that consumers perceive greater benefit to boycott participation
when they perceive a boycott situation as unjust as opposed to a situation where there is
little social injustice. Similarly, unless the boycott situation is perceived as unfair or
unjust, participating in an extended boycott of a preferred brand represents an
unnecessary cost to the consumer.
--------------------- -----
25
HE MEOther
MESelf
Moral
Emotions
BI BOYCOTT
Cost-Benefit
Evaluations
BENE
~~ COST
Note: HE = Humanitarian-Egalitarian Orientation; NAsH = Negative Attitude towards Big
Businesses; PMV = Perceived Moral Violation; MEOther = Other-condemning moral emotions;
MEselF Self-conscious moral emotions; BENE =Benefit evaluation; COST =Cost evaluation; BI =
Boycott Intentions; BOYCOTT= Boycott Behavior.
Figure 1. A two-process consumer boycott model.
Two distinct perceived benefits were identified in Klein et al.'s model.
Consumers will perceive greater benefit if they feel that they can make a difference and
have a positive impact on the situation or influence the company in some way by
boycotting. In addition, consumers will perceive greater benefit if they can enhance their
own self-image through boycotting. The greater the perceived benefit, the more likely
consumers will form pro-boycott intention.
Constrained consumption and counterarguments represent perceived costs to
consumers in the cost-benefit models. Doing without a preferred brand or product
introduces both psychological and utilitarian costs. Counterarguments suggest that
26
boycotting may actually do more harm (to others, to the self) than benefit. For example,
participation in the chocolate boycott may actually drive down global cacao bean prices
and force already poor cacao famlers to increase the use of slavery. Consistent with Klein
et al. (2004), perceived costs and boycott intention are inversely related. Consumers are
more likely to boycott when the perceived costs are low; conversely, consumers are less
likely to boycott when perceive personal costs are high.
Figure 1 also features two individual difference measures that are thought to
influence the perception of moral violation. A person with a higher humanitarian-
egalitarian orientation is likely to be more sensitive to moral violations. Katz and Irwin
(1988) suggested that people who hold strong humanitarian-egalitarian orientation are
concerned with equality, social justice, and others' wellbeing. In a boycott context,
people with strong humanitarian-egalitarian orientations are therefore likely to perceive
greater moral violation when a company's actions harm workers, consumers, or others.
Webster (1975) found that a negative attitude toward the power of big business was a
characteristic of socially conscious consumers. Thus, people who have a negative attitude
toward big businesses may also be more likely to judge a company's actions as being
unfair and perceive greater moral violation when confronted with a boycott situation.
Alternative Models
The sequencing of variables shown in Figure 1 assumes a parallel configuration in
which moral emotions and cost-benefit evaluations are each affected by perceived moral
evaluation, and both predict boycott intention. This hierarchy is simplified in Figure 2.
27
Percdved
Moral
Violation
Moral
Emotions
Cost-Benefit
Boycott
Intention
Figure 2. Hypothesized model.
However, it is also possible that alternative models exist that might better explain
the data. For example, one alternative is that emotions lead to a greater elaboration of
personal costs and benefits (see Figure 3). Kahneman's (2003) two-system model posits
that the rapid affect characterized by System 1 processing precedes the more deliberative
System 2 processing. Consistent with this perspective is Haidt's (2001, 2003) theory of
moral intuition, which suggests that people's initial response to a situation is affective
and these feelings are subsequently rationalized. Thus, the sequencing shown in Figure 3
suggests that the deliberation of costs and benefits may in itself be a form of emotional
coping (Duhachek, 2005; Duhachek & Iacobucci, 2005).
Perceived
Moral
Violation
Moral
Emotions
Cost-Benefit
Evaluation
Boycott
Intention
Figure 3. Affect-driven model.
------~~~ --~~
28
In contrast, it is also possible that consumers engage in greater evaluation of the
personal costs and benefits associated with a boycott when confronted with a perceived
moral violation. These deliberations would in turn elicit an emotional response that
would directly influence an intention to boycott. Such a sequence is shown in Figure 4.
I •Perceived Cost-Benefit A10ral BoycottMoral r--+ r---+
Violation Evaluation Emotions Intention
I t
Figure 4. Reasoning-driven model.
The three models featured in Figures 1, 2, and 3 by no means exhaust all of the
possibilities for alternative models presented by four variables. However, they do provide
a set of competing conceptualizations. Consistent with the goal of this dissertation,
comparing the perfonnance of each of these models will provide a deeper understanding
of how emotions and cost-benefit deliberations complement each other in consumer
boycott situations.
Counter-Messaging Tactics
The most common response to a consumer boycott is for a company to issue a
counter-message in the fonn of a press release. Klein et al. (2004) proposed that counter-
messaging could be effective in reducing consumer boycott intentions. The authors
recommended that finns can minimize consumer perceptions of moral violation by (a)
29
publicly acknowledging its action immediately, (b) providing a convincing explanation
for the firm's practices, and (c) publicizing the firm's ameliorating actions. Further, Klein
et al. prescribed three specific counter-messaging tactics to reduce the likelihood of
boycott participation:
1. The firm should acknowledge that it has heard the consumer's concern, explain its
actions, then stand firm on the decision to continue its course of action. The
objective of this type of message is to reduce the perception that consumers can
make a difference; thereby reducing the overall perceived benefit associated with
any consumer action.
2. Firms should convey the negative repercussions of boycotting. This tactic
provides counterarguments to consumers and seeks to minimize the ability for
consumers to feel better about themselves as a result of participating in boycott
activities. The objective of this tactic is to increase the perceived costs associated
with a boycott and to reduce the perceived benefits of such an action.
3. Boycotted firms should continue to promote the positive aspects of their product
in an effort to increase the perceived costs associated with a boycott.
The effectiveness of these counter-messages has not been empirically tested. The
hypothesized model suggests that it is likely that counter-messaging that addresses the
initial consumer perception of moral violation may be more effective than messaging that
subsequently addresses individual costs and benefits. Perceived moral violation is a
common antecedent to both emotional and deliberative processes. Reducing overall
moral violation should also reduce the impact of moral emotions and cost-benefit factors
30
on boycott intention. Thus, corporate public relations should focus first and foremost on
managing perceived moral violation in the media. Study 3 will test the effectiveness of
the counter-messages based on Klein et al.'s (2004) recommendations.
31
CHAPTER III
METHODOLOGY
Overview
Very little consumer boycott research has explicitly measured perceived moral
violation and emotions in boycott situations. As a result, the majority of consumer
boycott models have either omitted emotions entirely or addressed them obliquely. Thus,
a goal of this dissertation was to identifY and measure emotions that contribute to the
formation of a boycott intention. The primary variables in the hypothesized model were
perceived moral violation, moral emotions, cost-benefit factors, and boycott intention.
Two individual difference variables were also measured across all studies: humanitarian-
egalitarian orientation and negative attitude toward big businesses. Path analysis was the
primary tool used to examine the hypothesized relationships between variables, to assess
model fit, and to test alternative models. Data that included both emotions and cost-
benefit factors were modeled simultaneously to demonstrate a parallel two-process
model. Model variables and a description ofthe path analyses that were conducted are
discussed in detail below.
32
Key Model Variables
The variables measured across all three studies were perceived moral violations,
boycott intention, moral emotions, and cost-benefit factors. Each of these measures will
now be described.
Appraisal of Perceived Moral Violation
Perceived moral violation was measured directly after respondents' exposure to
the boycott stimuli using a scale created from Montada and Schneider's (1989) research
on justice and emotional reactions to the disadvantaged and Mauro et al. 's (1992) items
for legitimacy. Table 1 presents the four-item measure that uses a 7-point Likert scale to
state disagreement and agreement (1 = disagree, 7 = strongly agree).
Table 1. Perceived Moral Violation Measure
Variable
Perceived
Moral
Violation
(PMV)
1
2
3
4
Items
This situation is unfair for those affected.
I consider this issue a social injustice.
I believe the company's actions are egregious.
I regard the company's actions as unethical.
Boycott Intention
Boycott intention represents a spectrum of activities that reflect an escalating path
of militancy (Friedman, 1999; N. C. Smith, 2005). Table 2 presents the measure for
boycott intention. Each action indicator is a single-item scale using a 7-point Likert scale
(1 = not likely, 7 = very likely).
33
Table 2. Boycott Intention Measure
Variable
Boycott
Intention
(BI)
Items
I Discusses my dissatisfaction with regards to the company in
private.
2 Consider and think about withholding purchase of the
company's products.
3 Actually withhold purchasing the company's products.
4 Seek out and purchase a competitor's products.
5 Publicly discuss my dissatisfaction with the company.
6 Publicly discuss boycott and attempt to persuade other people to
Jom.
Moral Emotions
Table 3 presents a list of moral emotions (Haidt, 2001). Each of the emotions was
measured using a single-item 7-point Likert scale. Participants were asked to rate the
degree they felt an emotion in reaction to the boycott situation (l = did notfeel at all, 7 =
filt very strongly). The measure for other-condemning moral emotion (MEOther) used in
the analysis was created by calculating the mean of three moral emotions: anger,
contempt, and disgust. The self-conscious moral emotion (MEself) construct was
calculated as the mean of guilt, regret, shame, and embarrassment.
Table 3. Moral Emotions Measure
Variable
Other-condemning
Moral Emotion
(MEOther)
Self-Conscious
Moral Emotion
(MEse1f)
I
2
3
I
2
3
4
Moral Emotions
Anger
Contempt
Disgust
Guilt
Regret
Shame
Embarrassment
34
Consumer boycotts are closely associated with prosocial and helping behavior
(Klein, et al., 2004; Sen, et al., 2001) and customer retaliation (Gregoire & Fisher, 2007).
To build a more comprehensive list of emotions associated with consumer boycotts,
related prosocial (hope, empathy, and sympathy) and retaliation (betrayal, trust, hatred,
and suspicion) emotions were also measured. The presence of these emotions provides
further evidence of the role emotions play in consumer boycott. Further, these emotions
represent future research opportunities.
Cost-Benefit Factors
Cost-benefit factors represent deliberative assessments of perceived benefits and
costs associated with boycotting. Klein et al. ' s (2004) cost and benefit factors each had
two dimensions. Sub-scales for counterarguments and constrained consumption
represented perceived costs. Sub-scales for making a difference and self-enhancement
represented perceived benefits. Klein et al.'s (2004) measures were used verbatim except
for three items that were specific to the boycott context used in their study. The fourth
counterargument item ("I don't boycott [company] because it is a [country] company and
boycotting would lead me to buy foreign products") was replaced with a statement
regarding non-productive negotiation with the company, another type of counterargument
proposed by Friedman (1999). The constrained consumption items were adapted in each
study to be product and brand specific to the boycott stimuli. Each subscale is a multiple-
item construct rated on a 10-point Likert scale (1 = strongly disagree, 10 = strongly
agree). Table 4 summarizes the scale items used in this dissertation. For purposes of the
analysis, the benefit factor (BENE) measure was created by calculating the mean of the
items make a difference and self-enhancement. The cost factor (COST) measure was
created by calculating the mean of counterargument and constrained consumption.
35
Table 4. Cost-Benefit Factors
Variables Items
Benefit Make a 1 Boycotts are an effective means to make a
(BENE) Difference (MD) company change its actions
2 Everyone should take part in the boycott because
every contribution, no matter how small, is
important
3 By boycotting, I can help change [company's]
decision
Self-enhancement 1 I would feel guilty if! bought [company's]
(SE) products
2 I would feel uncomfortable if other people who
are boycotting saw me purchasing or consuming
[company] products
3 My friends/family are encouraging me to boycott
[company]
4 I will feel better about myself if I boycott
[company]
Cost Counterarguments 1 I do not need to boycott [company]; enough other
(COST) (CA) people are doing so.
2 I do not buy enough [company] products for it to
be worthwhile boycotting; it would not even be
noticed.
3 Boycotting may put this company's workers in
danger or cause unforeseen harm to those who are
not responsible for the situation.
4 Boycotting may be counterproductive to
negotiating an agreeable compromise with this
company.
Constrained 1 It would be difficult for me to give up [product]
Consumption 2 It would be difficult for me to give up [product]
(CC) from [company]
36
Individual Difference Variables
Two individual difference variables are expected to have a significant relationship
with perceived moral violation in the hypothesized model: humanitarian-egalitarian
orientation (HE) and negative attitude towards big businesses (NABB). Humanitarian-
egalitarian orientation describes an individual's sensitivity to unfairness and injustice
(Katz & Irwin, 1988). Negative attitude toward big businesses is expected to influence a
consumer's perceptions of a company's action (Webster, 1975). Although the two
individual difference variables are related only to perceived moral violation in the
hypothesized model, the variables will be treated as covariates in the path analysis.
Therefore, both HE and NABB will be directed initially to each of the constructs in the
model.
Humanitarianism-Egalitarianism Orientation
Humanitarian-egalitarian orientation (HE) varies from person to person and
inherently colors a person's perception of what is fair and just. Katz and Irwin (1988)
described HE as the "adherence to the democratic ideals of equality, social justice, and
concern for the others' well-being" (p. 894). Accordingly, Katz and Irwin's research
suggested that individuals who hold a strong HE may be more sympathetic to the plight
of the underdog and more sensitive to moral violations. Table 5 presents the HE scale (p.
905, Katz & Irwin 1988), which is a multiple-item construct that includes 10 items rated
on a 6-point scale (l = strongly disagree, 6 = strongly agree).
Table 5. Humanitarian-Egalitarian Scale
Name Items
37
Humanitarian-
Egalitarian
Orientation
(HE)
1
2
3
4
5
6
7
8
9
10
One should be kind to all people.
One should find ways to help others less fortunate than oneself.
A person should be concerned about the well-being of others.
There should be equality for everyone--because we are all human
beings.
Those who are unable to provide for their basic needs should be
helped by others.
A good society is one in which people feel responsible for one
another.
Everyone should have an equal chance and an equal say in most
things.
Acting to protect the rights and interests of other members of the
community is a major obligation for all persons.
In dealing with criminals the courts should recognize that many are
victims of circumstances.
Prosperous nations have a moral obligation to share some of their
wealth with poor nations.
Negative Attitude Toward Big Businesses
The corporate social responsibility literature suggests that attitudes toward big
business can shape consumer behaviors (Webster, 1975). Because symbolic boycotts
typically target large multinational corporations, participants with a negative attitude
toward big business may be more likely to perceive egregiousness in big business actions
and be suspicious of a firm's communications in boycott situations. People with strong
negative attitudes toward big businesses may be more likely to blame the company,
perceive the company's actions to be disingenuous, and be more willing to protest the
company (Friedman, 1996; N. C. Smith, 1990). People with a negative perception of big
business may also be more likely to take actions against them. Using a 7-point Likert
38
scale (1 = do not agree at all, 7 = agree a great deaf), Webster (1975) provided a single-
item measure asking for the strength of agreement with the following statement: "Big
business has too much power in this country."
Model Testing and Alternative Models
Path analysis was used in Study 2 and Study 3 to evaluate the fit of the
hypothesized model to the data. In addition, the hypothesized model was compared to
alternative models to assess the directionality of relationships. Path analysis is a special
instance of structural equation modeling (SEM) in which only observed indicators are
used to examine the causal structure of the model. The primary advantage of SEM is that
it permits the simultaneous estimation of all path coefficients instead of using a series of
multiple regressions. In addition, many SEM software programs provide fit statistics to
facilitate model specification and comparison. AMOS 7.0 was used for the path analyses
conducted in Study 2 and Study 3.
It is considered good practice to identify alternative models a priori. Figure 5
illustrates the configuration of the alternative models to be tested. First, two diagnostic
models (Models 1 and 2) were used test the directional relationship of key variables. In
Modell, the placement of perceived moral violation (PMV) and boycott intentions (BI)
was reversed. Because of the anticipated high correlation between these two variables, it
is necessarj to assess whether differences in the direction of relationships in the
hypothesized model exist. Model 2 examines the possibility that moral emotions and
cost-benefit evaluations predict PMV. This is consistent with earlier research in which
moral judgments are made at the end rather than at the beginning of the process of
making a moral decision (Haidt, 2001).
Modd I Model 2
MEOther
MEsdf
BENE
COST
Model 3 Model 4
Note: PMV =Perceived Moral Violation; MEOther= Other-condemning moral
emotions; MESeU=Self-conscious moral emotions; BENE = Benefit evaluation; COST
= Cost evaluation; BI = Boycott Intentions; Boycott = Boycott Behavior
(dichotomous)
Figure 5. Alternative models.
39
40
In Models 3 and 5, consistent with the configurations of coping models
(Duhacheck, 2005; Haidt, 2001) and the two-system model (Kahneman, 2003), moral
emotions were conceptualized as preceding cost-benefit factors in the causal sequence. In
Models 4 and 6, cost-benefit factors were positioned before emotions to test for
directionality in the causal sequence (the reverse of Models 3 and 5).
Multiple statistics were used as criteria for assessing model fit. In addition to the
traditional chi-square ci) test, Hu & Bentler (1995) proposed general guidelines for good
fit to be the Comparative Fit Index (CFI > .95), root mean squared error of approximation
(RMSEA < .06), and standardized root mean squared residual (SRMR < .08). Models not
meeting these criteria were considered a poor fit to the data. For model comparison, the
Akaike Information Criterion (AIC; Akaike, 1987; Anderson, Burnham, & White, 1998)
was used to compare the fit of models relative to each other to determine which model
was better performing.
The Comparative Fit Index (CFI) is a normed index (range 0-1) that compares the
model fit between the hypothesized model and the null model while considering sample
size (Bentler, 1990). The higher the value (> .95), the better fit of the model to the data.
The root mean square error of approximation (RMSEA) takes into account the errors of
approximation and accounts for the decrease in fit of the hypothesized model and data
due to the addition of parameters (i.e., lack of fit per df) (Hu & Bentler, 1999). RMSEA
values < .06 generally indicate a good fit to data. The standardized root mean squared
residual (SRMR) is the difference between standardized observed and predicted
41
covariance (Hu & Bentler, 1999). Thus, a value of zero would indicate a perfect fit of
model to data. In general, values < .08 indicate a good fit ofthe model to data. Finally,
the Akaike Information Criterion (AIC) is used to indicate relative fit for different models
estimated from the same data set (Akaike, 1987). AIC values are not directly
interpretable. Instead, the model with the lower AIC value indicates a better relative fit
between alternative non-nested models tested with the data.
Study 1: Exploratory Online Study of Active Boycotters
Study 1 was a descriptive study of real world boycotters. A mix ofquantitative
and qualitative data was gathered in Study 1 to measure emotional and cost-benefit
variables in real boycott situations and to generate a descriptive profile of boycotters. The
survey instrument is presented in Appendix B. Boycott participants from recognized
campaigns were recruited from a special interest group on a well-known social network
site (Facebook) to complete an Internet survey. In previous research, the cost-benefit
variables were developed fully while measures of emotions were either omitted (i.e., Sen
et aI., 2001) or indirect (i.e., John & Klein, 2003; Klein et aI., 2004). Gathering emotional
measures in addition to cost-benefit measures provided information about which
emotions, if any, were pertinent to boycott situations.
The survey gathered measures for key variables (emotions and cost-benefit
factors), individual differences (i.e., humanitarian-egalitarian orientation), and attitudes
(i.e., toward company, toward big businesses). In addition, participants were asked to
identify their boycott motivation using John and Klein's (2003) proposed typology of
--------------------- ---
42
motivations for consumer boycotts. This typology included six categories: (a) perceived
effectiveness and the illusion of control, (b) altruism, (c) a thrill of victory, (d)
expressive, (e) punitive, and (f) clean hands.
The survey also asked respondents to elaborate on their participation through
three open-ended questions. First, respondents were asked how they would respond if the
boycott target conceded to boycotter demands. Second, they were asked how they
became involved in the boycott. Finally, respondents were asked specifically if they felt
that boycotting is a moral act and differentiates them from their peers, as described by
Kozinets and Handelman (1998, 2004).
The open-ended responses were coded for content analysis. First, the boycott
targets identified by participants were categorized into boycotts against a country, a firm
(MNC), a single brand, an entire product category, or an individual. Second, responses
regarding boycott recruitment were used to create categories for sources of boycott
recruitment. Four sources were identified: social groups and affiliations, peers, sponsored
announcements, and personal research. The open-ended question regarding what
boycotters would do if boycott target conceded to demands was coded simply to specify
whether the boycotter would end the boycott or continue to boycott. For the final open-
ended question regarding whether boycotting differentiated participants from their peers,
the answered were simply coded to confirm or disconfirm the moral differentiation from
peers.
43
Study 2: Conceptual Model Testing
The primary purpose of Study 2 was to examine the sequencing between
perceived moral violation, boycott intention, and boycott behavior. Klein et al. (2004)
found that perceived moral violation affected boycott intention (appraisal -7 intentions),
while Sen et al. (2001) provided empirical support for the effect of boycott intention on
actual boycott behavior (intention -7 behavior). Study 2 measures each of these variables
in the same experiment and explores their hierarchical ordering, along with moral
emotions and cost-benefit factors as outlined in the hypothesized model shown in Figure
1.
Data were collected for Study 2 in three separate stages to reduce demand and
carryover effects. In the first stage, in an unrelated study, respondents provided
information on their brand preferences and participated in an online shopping simulation.
In the online simulation, participants were presented with seven product categories as
part of a shopping list. The survey randomly presented each product category with two
dominant national brands. For each product category, participants were asked to choose
their preferred brand to include in their shopping basket to fulfill the requirements of the
shopping list-this choice was used to indicate the participant's preferred brand for a
product category. Respondents were also given a "no purchase" option.
The brand preference information from the first stage was used to randomly
assign student participants into one of three product boycotts in the second stage of the
study. The three product categories were selected to be common, frequently purchased,
and easily substitutable products. These are characteristic of products that are commonly
44
boycotted (Friedman, 1996). To help disguise the intent of the study, participants were
asked to provide their reaction to three types of articles presented in random order:
positive, negative (boycott), or incidental. Each of the articles had approximately the
same number of words. Positive and incidental articles provided positive or neutral
information about a single brand that did not involve a preference derived from the first
session's data collection. In contrast, boycott recruitment articles referred to the
participant's preferred brand from the first stage of data gathering.
Participants were sent customized invitations keyed to the brand they chose to put
in their shopping basket in the first stage. The key corresponded to the specific product
boycott condition (Table 6). For example, if a student was assigned into the battery
boycott condition, he or she would receive a link to the Condition 3 sequence and receive
the boycott recruitment stimuli specific to the brand chosen in the first portion of the
study. A summary of the order of articles shown to participants is provided in Table 6.
Table 6. Product Category Stimuli Presented to Participants
Toothpaste
Batteries
Boycott
(lVeuative)
Incidental
Positive
Positive
Boycott
(Negative)
Incidental
Positive
Boycott
(Negative)
The actual experimental manipulations appeared only for the boycott recruitment
article in the second stage of the study. Study 2 used a 3 (Cost-Benefit Perspective vs.
45
Emotional Perspective vs. Control) x 2 (Positive Framing vs. Negative Framing)
between-subjects design. Students (N = 201) participated in the study in exchange for
class credit. For the negative boycott recruitment article, participants were randomly
assigned to read one of six boycott stimuli for the brand they preferred. Sen et al. (2001)
manipulated pro-boycott message framing either to highlight the positive or negative
outcomes of the boycott. Participants were assigned to read a message framed in either
positive or negative terms regarding the outcome of boycott activity. Next, participants
assigned to the cost-benefit perspective were asked to read each article with a rational and
objective viewpoint, paying special attention to the costs and benefits to themselves (i.e.,
"read the article from an objective and rational perspective and consider the economic
impact on you"). Participants assigned to the emotional perspective were asked to read
the article, eschewing costs and benefits to themselves, and to attend to their feelings
(i.e., "read the article paying special attention to your feelings and how you feel about
the situation "). The final condition was a control condition and provided no prompt for
how to read the article. Measures for perceived moral violation, boycott intention,
emotions, and cost-benefit factors, as well as manipulation checks were gathered after
reading the boycott recruitment stimuli.
In the final data gathering session two to five days after the administration of the
boycott stimuli, participants were asked to complete a second online shopping simulation
similar to ~hat in the first stage to see whether purchase behaviors had changed. The three
product categories that were being boycotted were randomly shown with four new
46
product categories as part of the shopping list. The student participants were then
debriefed. All survey instruments and stimuli for Study 2 are presented in Appendix C.
Study 3: The Effect of Counter-Messaging on Boycott Intention
The purpose of Study 3 was to test the effectiveness of the four counter-message
options proposed by Klein et al. (2004) using a non-student sample. The endogenous
variable in this study was boycott intention (BI). The most effective counter-message
should demonstrate the greatest decrease in the BI measure. The study focused on a
single real-world boycott (Global Exchange chocolate boycott) and used a leading U.S.
chocolate company as the boycott target. A commercial consumer panel was recruited to
participate in Study 3 to improve external validity and demonstrate immediate market
application. The format of the study was an online survey.
Study 3 was a simple random group design. Participants were randomly assigned
to either a control condition that did not receive a counter-message or one of five counter-
message conditions (all stimuli for Study 3 are presented in Appendix D). All participants
were asked to read a pro-boycott message in the form of a press release issued by a
fictitious prosocial group. Participants reading one of the counter-messages received a
second press release after reading the first that contained one the following counter-
message tactics proposed by Klein et al. (2004) and written specifically for Study 3:
1. Acknowledge action and communicate positive changes. The tactic attempted to
reduce perceived moral violation by acknowledging wrongdoing and
communicating ameliorating changes and practices.
47
2. Reduce perception ofboycott efficacy. The second tactic attempted to reduce the
perception of boycott efficacy by standing firm on the company policy and
practices, and by creating the perception that consumer actions would not make a
difference, thereby reducing the overall benefit of boycotting.
3. Explain negative repercussions ofconsumer's boycott action. The third tactic
tried to increase counterarguments by elaborating on the negative repercussions of
boycotting. The overall effect sought was to increase perceived boycott cost
(counterargument) and decrease benefit (self-enhancement).
4. Emphasize the positive aspects ofthe brand andproduct. The tactic attempted to
increase constrained consumption by promoting the positive benefits of the
product and the brand.
The actual response from the offending company was also included as a condition.
5. Denial and communicate leadership. Rather than acknowledge wrongdoing, the
original response denied allegations of slaves used for harvesting cacao. The press
release provided counter-factual evidence and provided statements of positive
company action. The original message essentially attempted to divert attention
away from the claims of slave labor and reframe the situation to put the company
in a favorable position.
Participants were asked to complete a counter-message manipulation check and
measurements of key dependent variables immediately after being shown the stimuli. The
survey instrument is presented in Appendix D.
48
CHAPTER IV
RESULTS
Study 1
Sample Characteristics
Individualized survey invitations were sent electronically to 400 people, selected
from a boycott interest group on a well-known social networking site. The interest group
had approximately 4,000 members. Participants were screened based on two criteria.
First, they had to have posted one or more comments to the interest group forum. Second,
they had to have identified a specific boycott target in their posting. The response rate of
usable online surveys was 30.2% (N = 121). This response rate was similar to rates
reported in previous Internet survey research in marketing and advertising (i.e., 29.7% in
Morrison & Haley, 2006; 29% in Reast, Palihawadana, & Shabbir, 2008).
The sample characteristics for all studies are presented together in Table 7 for
comparison purposes. The typical online boycotter in the sample appeared to be
Caucasian, college-educated men and women in their early twenties with positive
attitudes toward boycotting and negative opinions of large businesses. In addition, they
held strong humanitarian-egalitarian viewpoints (see Table 8). Although unions have
traditionally been very active in organizing boycotts, very few ofthe boycotters belonged
to a union (n = 16, 13.2%). The general description of modern boycotters provided by the
sample was consistent with findings in earlier boycott research (Friedman, 1999; N. C.
Smith, 1990). Specifically, boycotters tended to be socially conscious consumers who
were white, pre-middle aged adults of moderate-to-high socio-economic status O'J. C.
Smith, 1990, p. 178).
49
Table 7. Sample Characteristics (Studies 1-3)
Study 1 Study 2 Study 3
(N = 121) (N = 201) (N = 709)
Demographics n (%) n (%) n (%)
Sex (Female) 60 (49.6%) 84 (41.8%) 338 (47.7%)
Age
16-19 12 (9.9%) 5 (2.5%) 18 (2.5%)
20-24 88 (72.7%) 184 (91.5%) 126 (17.8%)
25-34 21 (t7.4%) 12 (6.0%) 136 (19.2%)
35-44 181 (25.5%)
45-54 194 (27.4%)
55+ 54 (7.6%)
Race
White/Caucasian 88 (72.7%) 157 (78.1%) 575 (81.1%)
Middle Eastern 22 (18.2%) t2 (1.7%)
Hispanic 32 (4.5%)
African!African American 9 (7.4%) 2 (1.0%) 35 (4.9%)
Asian/Pacific Islander 5 (4.1%) 27 (13.4%) 49 (6.9%)
Other 7 (5.8%) 10 (5.0%) 6 (0.8%)
Education
Primary 10 (1.4%)
Secondary 12 (9.9%) 112 (15.8%)
Tertiary (College/University) 86 (71.1%) 176 (87.6%) 471 (66.4%)
Graduate/Professional 23 (19.0%) 25 (12.4%) 116 (16.4%)
Union Member 16 (13.2%) 167 (23.6%)
50
Table 8. Individual Differences (Study 1)
Scale M SD
Attitude
Big Business (NABB) 1-7 5.43 1.13
Boycott 1-7 6.08 1.38
Unions 1-10 7.83 1.46
Humanitarian-Egalitarian Orientation (HE) 1-6 5.35 0.32
Individual Boycott Participation
Three open-ended questions were asked to ascertain how respondents were
recruited to participate in a boycott, their willingness to end the boycott, and if they felt
their actions differentiated them from peers. For each of the questions, two student
assistants coded the responses for analysis based on the categories and labels established
by the primary researcher. Inter-rater agreement was .96 (Rust & Cooil, 1994); the
primary researcher used personal judgment to resolve coding differences for analysis.
As shown in Figure 6, the common ways that participants became aware of
potential boycotts were through social groups and affiliations (n = 61. 50.4%), peers (n =
40,33.1 %), sponsored and public announcements (n = 17, 14.0%), and personal research
(n = 3, 2.5%). Social groups and affiliation included any organized groups and
affiliations a person belonged to, including unions, clubs, political organizations, online
groups, etc. For many, individual peer-to-peer conversations provided much ofthe
background for recruitment and participation in boycotts. Advertisements and public
service announcements from unaffiliated groups provided the point of recruitment for the
------- -------
51
remaining participants. For personal research, a few participants noted they initiated their
own boycotts through investigative research and shared their findings with others.
70
In 50.4%
....
= 60C'Il
c..
....
~
'J:l 50l.
C'Il
~
:t 40
0
~ 300CQ
....
0 14.0%l-o 20
ClI I,.QE 10= 2.5%z 0
--
Interest Groups Direct Peer Sponsored Personal
Announcements Research
Figure 6. Source of boycott recruitment.
Figure 7 illustrates how boycotters would react if the boycott target conceded to
boycotter demands. Most boycott participants (n = 57, 47.1 %) stated that as long as the
boycott target made suitable reparations to the affected parties (and put in place processes
to end the egregious acts), they would be willing to end the boycott. However, a large
number of boycott participants were unwilling to end their boycott even if the target
remedied the situation (n = 49,40.5%). Of those unwilling to end the boycott, eight
(6.6%) noted that they would continue to boycott with the goal of "putting the company
out of business because there is no recourse for their immoral actions." So, worse than
simple exit behavior, the initial act of boycotting contributed to continued activism
52
against a target. About 12.4% (n = 15) of boycotters were uncertain or did not specify
whether they would end their individual boycott. Almost all participants had already
found alternatives and were happy with the alternatives or were able to do without
products from the boycott target (n = 115,95.0%).
60
~
== 50te
l:l....
.~
....
I-. 40~
c..
:t::
0 30~
>.
0
CQ
.....
200
I-.(lJ
,.Q
6 10
=:z
0 ----
End Boycott Continue Boycott Uncertain/Unknown
Figure 7. Likelihood of ending boycott if target concedes to demands.
Only 59 (48.8%) respondents provided information on the final open-ended
question regarding whether their boycott actions differentiated them from their peers.
Four identified their actions as differentiating themselves from their peers, as suggested
by Kozinets and Handelman (1998). The remainder only sought differentiation from
people outside the peer or affiliation group that recruited them into the boycott. Boycott
participants felt more alike or connected with people in their boycott group and felt
rewarded by their likeness or solidarity. For example, one respondent noted, "Through
53
my actions, I want to demonstrate that my friends and I are more responsible than the
average consumer who is pretty much clueless."
The results of the open-ended responses suggested that current boycotts are an
action of both first and last resort that is spread primarily by peers and word of mouth.
Although boycotting may be a form of individual moral expression, attention to a target's
egregious act is often guided by group affiliations (e.g., political party). Further, many
boycotters continue to boycott as a punitive measure even if the target takes corrective
action. The implication is that practitioners must handle boycotters delicately. Boycotted
firms cannot assume remedies will automatically end boycott actions and repair negative
attitude and feelings toward the firm, and must woo boycotters back as new customers
again.
Boycott Targets and Motivations
The reported boycott targets were coded into six categories, defined as boycotts
against (a) a country, (b) a multinational corporation (MNC) or global firm, (c) a single
brand, (d) a single product category or industry, or (e) an individual. Each participant was
also asked to self-select the primary motivation from a choice of six motivations: (a)
control/effectiveness, (b) altruism, (c) thrill of victory, (d) expressive, (e) punitive, and (f)
clean hands. The six motivations were further grouped into instrumental (controll
effectiveness, altruism, and the thrill of victory) and symbolic (expressive, punitive, and
clean hands) boycotts (John & Klein, 2003). Table 9 summarizes respondent boycott
motivations and the target of the boycott.
[ ,
54
Table 9. Boycott Targets and Motivations
Boycott Target
Single Product
Motivation Country Firm (MNC) Brand Category Individual Total
Instrumental
Control/Effectiveness 7 1 8
(5.8%) (.8%) (6.6%)
Altruism 1 21 2 1 25
(.8%) (17.4%) (1.7 %) (.8%) (20.7%)
Thrill of Victory 2 2
(1.7%) (1.7 %)
Symbolic
Expressive 20 1 3 24
(16.5%) (.8%) (2.5%) (19.8%)
Punitive 5 25 7 2 39
(4.1%) (20.7%) (5.8%) (1.7%) (32.2%)
Clean Hands 8 8 6 1 23
(6.6%) (6.6%) (5.0%) (.8%) (19.0%)
14 83 16 6 2 121
Total (11.6%) (68.6%) (13.2%) (5.0%) (1.7%) (100%)
Note: l(20) = 41.7, p < .01.
The frequency of primary motivation for boycotting across target categories is
consistent with prior research. Boycotts against large MNCs and global firms (for
example, Coca-Cola, Nestle, Unilever) accounted for greater than two thirds of the
reported boycott targets (n = 83, 68.6%). It appeared that symbolic motivations (n = 86,
71.1%) outnumbered instrumental motivations (n = 35, 28.9%). These results were
consistent with assertions by Friedman (1999) and Kozinets and Handelman (1998) that
modern boycotts were more often symbolic than instrumental. The primary motivation
for instrumental boycotts appeared to be altruism (n = 25,20.7%), which was consistent
with the notion that modem boycotts are often prasocial in nature (i.e. Sen et aI., 2001;
Smith, Klein, & John; 2004). The most frequently cited motivation for symbolic boycotts
--------
55
was punitive (n = 39,32.2%), followed by expressive (n = 24,19.8%) and clean hands (n
= 23, 19.0%).
The frequency of altruism, expressive, punitive, and clean hands as primary
motives is suggestive that boycott intention is the result of an emotional process.
Altruism has often been linked to coping behaviors related to empathy and sympathy
(Batson, 1998; Eisenberg et aI., 1989), moral judgment and empathic emotions (Hoffman,
1987), and moral emotions (Haidt, 2003). The punitive and clean hand motivations are
closely associated with other-condemning and self-conscious moral emotions (Haidt,
2003). Therefore, emotional expression may be as much a part of boycotting as cost-
benefit factors.
Cost-Benefit and Emotional Characteristics
Comparisons were made between instrumental and symbolic boycotts to identify
basic patterns for key emotional and cost-benefit variables. These variables included
multiple facets of perceived moral violation, emotions, and cost-benefit factors. The one-
sample Kolmogorov-Smirnov (K-S) test was selected to make comparisons because
initial inspection of data indicated non-normal distributions and unequal group sizes. The
K-S test is a non-parametric test used to see whether two data sets differ significantly; the
test makes neither assumption about the distribution of the data nor sizes of independent
data sets. The following analysis used the K-S test in two ways. First, the K-S test
determined whether means differed significantly from scale midpoints to qualitatively
assess "high" and "low" patterns for key variables in the reported boycotts. Second, the
56
K-S test was used to compare means between instrumental (n = 45) and symbolic (n =
86) boycott groups to examine if there were patterns of differences. Table 10 presents the
summary of two sets of tests.
All means except for counterargument were significantly different from the scale
midpoint (all ps < .05). The means for embarrassment (M = 1.58, SD = .62) and shame
(M= 2.38, SD = 1.31) were low compared to guilt and regret. Benefits appeared to be
evaluated to a greater extent than costs in the sample. Both make a difference (M = 6.77,
SD = 1.47), and self-enhancement (M = 6.86, SD = 1.57), P < .05, appeared to have a
greater role than cost factors. Counterarguments did not appear to have figured greatly
into the reported boycotts (p> .05) and the mean for constrained consumption was low
(M= 3.75, SD = 1.08). However, it is likely that most boycotts were for easily
substitutable goods and services and the reason constrained consumption might have
been low relative to other factors.
The comparisons between instrumental and symbolic boycotts yielded notable
results. The importance of perceived moral violation and cost-benefit evaluations were
consistent across both types of boycotts and little difference was found (all ps > .05);
however, significant differences were found for other-condemning emotions, prosocial
emotions, betrayal, suspicion, and hatred (all ps < .05). The patterns suggest that higher
means for other-condemning emotions and prosocial emotions differentiate symbolic
from instrumental boycotts. It appeared that emotions are a necessary component for
symbolic boycotts whereas cost-benefit factors alone suffice for explaining boycott
intention in traditional instrumental boycotts.
-------------------
Table 10. Differences in Key Measures for Instrumental vs. Symbolic Boycotts
57
Compared to Scale Midpoint InstlUmental vs.
Combined InstlUmental Symbolic Symbolic
M (S.D.) M (SD.) M (SD.) K-S (5Ng.)
Percieved Moral Violation (7-point Likelt scale)
Injustice 5.91 (1.02) 5.60 (.98) 6.03 (1.01) ns
Unfair 6.52 (.50) 6.43 (.50) 6.56 (.50) ns
Unethical 6.06 (.83) 5.91 (.74) 6.12 (.86) ns
Egregious 6.30 (.72) 6.29 (.75) 6.30 (.70) ns
Moral Emotions (7-point Likert scale)
Other-condemning
Anger 5.93 (1.07) 5.29 (1.20) 6.20 (.89) ***
Contempt 5.77 (1.09) 5.37 (.94) 5.93 (l.12) **
Disgust 4.01 (1.65) 2.89 (2.01) 4.47 (l.22) ***
Self-Conscious
Embarassment 1.58 (.62) 1.46 (.51) 1.63 (.65) ns
Shame 2.38 (1.31) 2.17(.92) 2.47 (1.42) ns
Guilt 5.75 (1.41) 5.40 (1.93) 5.90 (Lll) ns
Regret 4.85 (1.57) 4.91 (1.79) 4.82 (l.48) ns
Cost-Benefit (1 O-point Likert scale)
Benefit
Make a difference 6.77 (1.49) 6.97 (1.51) 6.68 (1.49) ns
Self-enhancement 6.86 (Ll4) 6.78 (l.48) 6.89 (.97) ns
Cost
Counterargument 5.09 (1.01)'" 5.240.18)/1·, 5.03 (.94)'" ns
Constrained consumption 3.75 (1.08) 3.89 (LlO) 3.69 (1.07) ns
Related Prosocia1 Emotions (7-point Likert scale)
Hope 4.69 (1.27) 4.09 (Ll2) 4.94 (1.25) *
Empathy 5.41 (1.38) 4.60 (2.08) 5.73 (.76) *
Sympathy 5.70 (1.14) 5.14 (1.73) 5.93 (.66) **
Related Feelings (7-point Likert scale)
Betrayal 5.02 (1.27) 4.37 (1.19) 5.28 (1.20)
Suspicion 4.55 (1.71) 4.09 (2.02)",1 4.74 (1.54)
TlUSt 1.97 (.87) 2.14 (.77) 1.90 (.90)
Hatred 5.03 (1.67) 4.43 (2.4) 5.28 (1.17)
Note: All comparisons to scale midpoint significant p < .05 unless specified.
.01, *** P < .001) ns = non-significant.
**
*
ns
***
* p < .05, ** P <
58
Anger and contempt were the two most prominent other-condemning moral
emotions reported. Anger was higher in symbolic (M = 6.20, SD = .89) than in
instrumental boycotts (M = 5.29, SD = 1.20), p < .001. This was consistent with the
majority of prior literature that lists moral outrage as a foundation of consumer boycotts
(i.e., Friedman, N. C. Smith). Contempt was also higher for symbolic (M= 5.93, SD =
1.12) than for instrumental boycotts (M = 5.77, SD = 1.09), p < .01. The importance of
disgust was unclear and may require additional research, as it had less of an effect in
instrumental boycotts (M= 2.89, SD = 2.01) than in symbolic boycotts (M= 4.47, SD =
1.22).
The data suggested that prosocial emotions also mattered in boycott situations.
The conceptualization of boycotts as being driven by altruism is supported by the high
ratings for prosocial emotions such as hope, empathy, and sympathy scores (Ms = 4.69,
5.41,5.70, SDs = 1.27, 1.38, 1.14, respectively). In addition, hope was higher for
symbolic (M= 4.94, SD = 1.25) than instrumental boycotts (M= 4.09, SD = 1.12),p <
.05. Further, empathy and sympathy were greater for symbolic (M= 5.73, 5.93, SDs =
.76, .66) compared to instrumental groups (M= 4.60,5.14, SDs = 2.08, 1.73),ps < .05,
.01.
Emotions such as betrayal, suspicion, trust, and hatred as outlined in the corporate
social responsibility literature were also relevant for boycotting behavior (Gregoire &
Fisher, 2007). Although these emotions were not the central focus of the present research,
they help form a more global list of boycott-related emotions and represent potential
future research topics. The emotion of betrayal was felt more keenly by symbolic (M =
59
5.28, SD = 1.20) than instrumental boycotters (M= 4.27, SD = 1.19),p < .01. Similarly,
the visceral emotion of hatred was more acute in symbolic (M = 5.28, SD = 1.17) than
instrumental boycotts (M = 4.43, SD = 2.40), P < .001. Finally, boycotters that cited
symbolic motivation tended to be more suspicious (M = 4.74, SD = 1.54) than those with
instrumental motivations (M = 4.09, SD = 2.02),p < .05. Unsurprisingly, distrust was
uniform between groups (Ms = 2.14, 1.90, SDs = .77, .90),p < .05. These additional
negative emotions further support the notion that many consumer boycotts involve
emotion and should therefore be considered individually in future research.
Summary
The exploratory study suggested that an emotional process was central to
consumer boycotts in three ways. First, although affective motives have been mentioned
in prior research (Friedman, 1999; Klein et aI., 2004; Sen et aI., 2001; N. C. Smith,
1990), this was the first study to measure and survey how multiple discrete emotions
differed depending on boycotting motives. Notably, symbolic boycotts-expressive,
punitive, and clean hand motivated boycotts-were differentiable emotionally from
instrumental boycotts. In addition, the frequency of altruistic motives, along with the
importance ofprosocial emotions, supported the notion that boycotting is a helping
behavior. Finally, perceived moral violation was a critical component of both symbolic
and instrumental boycotts.
Second, the comparison between symbolic and instrumental boycotts was not
intended to specifically test for differences between the two types of boycott motivations,
60
but, rather, to demonstrate that variability exists for moral and prosocial emotions in
different boycott situations. The results suggested the existence of variables that have
been ignored in earlier research. For example, participants in symbolic (vs. instrumental)
boycotts demonstrated higher other-condemning and prosocial emotions. Therefore, it is
important to include emotions as a distinct and separate contributor to the boycotting
phenomenon. Extensive research has already provided linkages between empathy and
sympathy to prosocial behavior (e.g., Batson, 1998; Madrigal, LaBarge, & Chen, 2007 ),
so moral emotions were the primary focus of the next two studies.
Third, this study established a list of emotions that can be used for future boycott
research. The relationship between corporate social responsibility and betrayal, trust,
hatred, and suspicion (see Gregoire & Fisher, 2007) provides fertile ground for future
work but is well beyond the scope of the current dissertation. Further, a list of boycott-
related emotions might provide practitioners with insights into a greater variety of tactics
and options for pro- and counter-boycott messaging and promotions.
Study 2
Study 2 attempted to establish the relationship between perceived moral violation,
boycott intention, and boycott behavior. A series of logistic and linear regressions tested
the basic assumption that perceived moral violation was a predictor for boycott intention
and behavior, and boycott intention was a predictor of boycott behavior. A mediation test
was used to determine whether boycott intention mediated the effects of perceived moral
violation on boycott behavior. In addition, Study 2 attempted to provide initial support
61
that consumer boycott decisions have both emotional and cost-benefit dimensions. To
demonstrate that boycotting can be conceptualized as a consumer coping behavior, path
analyses were used to examine the relationship between perceived moral violation (PMV)
and boycott intention (B1) by incorporating moral emotions and cost-benefit factors as
intervening variables. A coping behavior pattern was established by demonstrating the
following path (Tangney et aI., 2007): appraisal (perceived moral violation) -7 emotion
-7 intentions (boycott intention) -7 coping behavior (boycott behavior). Finally, a
secondary path analysis attempted to identifY individual discrete emotions and cost-
benefit factors that contributed directly to boycott intention and behavior.
Sample Characteristics
A total of201 student participants provided complete and usable responses to a
three-part online survey. Eligible participants had to complete all three surveys and pass
manipulation checks that verified the product category and boycott target brand name at
the end of the second survey. The online survey randomly assigned participants into 1 of
12 conditions based on perspective, framing, and boycott stimuli. Table 7 (previous
section) provides the combined summary of sample characteristics. A student sample
was selected because the demographic characteristics were similar to the group of active
participants in Study 1 for race (Caucasian = 157, 78.1%),education(M= 15.3,SD=
1.70), and age (the median age was 21).
62
Descriptive Statistics
Table 11 summarizes the descriptive statistics and provides the correlation matrix
of the variables in the model. Data for different product categories were combined to
generalize across product categories. The aim of the manipulations was to create variance
in the data for path analysis using structural equation modeling (Brown & Dacin, 1997).
The outcome variable was whether the respondent intended to withhold consumption
during the shopping experience, referred to here as boycott behavior (BOYCOTT).
Consistent with Sen et al.'s (2001) second study, the variable was coded as potential
"boycott" (1) if the participant changed brands or withheld purchase, and "no boycott"
(0) ifthe participant continued to purchase the same brand. In addition, the data were
centered by subtracting the variable mean from each variable to attempt to minimize
multicollinearity issues. Because several variables are hypothesized to be mediators in the
model, multicollinearity may be problematic in regressions for variables that both directly
and indirectly influence a dependent variable. Finally, the experimental manipulations
were dummy coded along with interaction terms for inclusion into the analysis to account
for the effects of experimental manipulations. With the exception of individual difference
variables, correlations between all variables in the model were significant (all ps < .05)
and consistent with the valence suggested by the hypothesized model (see Table 11).
Both humanitarian-egalitarian orientation (HE) and negative attitude toward large
corporations (NAss) were significantly correlated with perceived moral violation (PMV),
r = .32, .23, respectively, bothps < .01.
Table 11. Descriptive Statistics and Correlation Matrix (Study 2)
Behavior Intention Appraisal Moral Emotions Evaluations Individual D(fferences
BOYCOTT BI PMV MEOthcr MEscH' BENE COST HE NA BB
BOYCOTT I
BI .51 **
PMV .32** .59** 1
MEOthcr .31 ** .54** .59**
MEsclf .20** .31 ** ,32** .50** 1
BENE .44** .68** .57** .52** .30** 1
COST -.14** -.24** -.21 ** -.23** -.15* -.32** 1
HE .22** .32** .38** .28** .19** .33** -.04 1
NA BB .08 .23** .26** .10 .11 .16* -.09 .27** I
n 201 201 201 201 201 201 201 201
M 3.72 5.57 4.21 3.20 5.50 4.04 4.60 4.81
(SO) (1.54) (1.11) (1.27) (1.40) (2.21 ) (0.85) (0.68) (1.51)
# Items 1 6 4 3 4 7 6 10 1
a -- 0.88 0.89 0.7 0.84 0.88 0.73 0.88
Notes:* p < .05; ** p < 0.01 level; BOYCOTT = Boycott shopping behavior; BI = Boycott Intent; PMV =
Perceived Moral Violation; MEOthcr = Other-condemning moral emotions (Anger, Contempt, Disgust); MEseir =
Self-conscious Moral Emotions (Guilt, Regret, Shame, Embarrassment); BENE = Benefit Evaluation (Make a
differences, Self-enhancement); COST = Cost Evaluation (Counter-argument, Constrained Consumption); HE =
Humanitarian-Egalitarian Orientation; NAnn = Negative Attitude towards Big Businesses.
0'.
w
t-----------
64
An initial comparison based on the BOYCOTT dichotomy (non-boycotters vs.
boycotters) using t-tests indicated that boycott participants were more likely to exhibit
greater boycott intention (BI), perceived moral violation (PMV), and other-condemning
(MEOther) and self-conscious (MESel£) moral emotions (see Table 12). BI was higher
for potential boycotters (M = 4.57, SD = 1.31) than non-boycotters (M = 2.99, SD =
1.33), t(199) = 8.43, p < .001. Similarly, PMV was greater for potential boycotters (M =
5.95, SD = 1.09) than non-boycotters (M = 5.24, SD = 1.03), t(199) = 4.79, p < .001. In
addition, the values for moral emotions were higher for potential boycotters (MMEOther
= 4.65, MMESelf= 3.50, SDs = 1.12, 1.36, respectively) than for non-boycotters
(MMEOther = 3.85, MMESelf= 2.95, SDs = 1.28, 1.39, respectively), t(199) = 4.65,
2.82, ps < .001, .01.
Table 12. Means Comparisons for Non-Boycotters and Boycotters (BOYCOTT)
BOYCOTT
No Boycott Boycott
(n = 109) (n = 92)
M Sf) }~f Sf)
Perceived Moral Violation 5.24 1.03 5.95 1.09
Boycott Intentions 2.99 1.33 4.57 1.31
4.79***
8.43***
Moral Emotions
Other-Condemning
Self-Conscious
Cost-Benefit Factors
Benefit
Cost
3.85 1.28
2.95 1.39
4.61 2.07
4.15 1.92
4.65 1.12
3.5 1.36
6.55 1.9
3.91 1.74
4.65***
2.82**
-6.87***
-2.07*
Individual Differences
HE 4.70 1.63
NA ml 4.46 1.69
Note: * p < .05, ** p < .01; *** P < .001
4.95 1.35
4.76 1.64
3.21 **
1.13
65
The results were also consistent with prior cost-benefit literature. People who
boycotted (M = 6.55, SD = 1.90) judged the overall benefit to be greater than non-
boycotters (M= 4.61, SD = 2.07), t(l99) = 6.87,p < .001; conversely, it was not
surprising that non-boycotters perceived greater cost (M = 4.70, SD = 4.15) than
boycotters (M= 3.91, SD = .74), t(l99) = 2.07,p < .05.
Finally, people that held stronger humanitarian-egalitarian orientations were more
likely to boycott. The individual difference rating for boycotters (M = 4.95, SD = 1.35)
was significantly higher than non-boycotters (M = 4.70, SD = 1.63), t(l99) = 3.21, P <
.001. However, no significant differences were found for attitude toward big businesses,
t(199) = 1.13, ns.
Appraisal, Intention, and Behavior
The proposed model (see Figure 1) suggested that there is a direct relationship
between perceived moral violation (PMV), boycott intention (BI), and boycott behavior
(BOYCOTT). Logistic and linear regressions ascertained ifBI was a strong predictor of
BOYCOTT and ifBI mediated the influence of PMV on BOYCOTT. This relationship
was important to establish before adding additional process variables between PMV and
BI because the BI scale used in Study 2 is different than those used by Sen et al. (2001)
and Klein et al. (2004). It was important to demonstrate that the BI variable in the current
study behaved in a similar manner compared to previous studies, and that PMV, BI, and
BOYCOTT were measured together in the same study. Table 13 provides the results of
logistic and linear regressions and Table 14 provides the associated fit statistics.
----------_._-----~
66
Table 13. Logistic Regression for Perceived Moral Violation and Boycott Intent as
Predictors of Boycott Behavior
Model: Predictors B SE Wald c{j' hP e
Mod,,)] 1
Constant -.2 J .15 1.92 I .17 .81
PMV .67 .15 18.89 1 <: .001 1.95
Model 2
Constant -.25 .17 2.28 .13 .78
HI .87 .14 41.05 <: .001 2.38
Model 3
Constant -.25 .17 2.30 .13 .78
PMV .08 .] 9 .]8 .67 ] .08
HI .83 .16 28.85 <: .OOJ 2.30
Table 14. Fit Statistics for Boycott Behavior Models
Model summary Goodness-of-fit
Model Predicted df Likelihood Cox & Nagelkerke df
Correct X' p ratio Snell R) R) X' P
71.1%1 22.3 I <.001 254.9 .105 .140 19.6 8 0.01
2 71.J!% 59.6 J <.OOJ 217.6 .257 .343 6.48 8 0.59
.. 70.6% 59.8 2 <.01 217.4 .257 .344 5.5 8 0.7],)
PMV was a significant predictor of whether someone chose to boycott or not
(BOYCOTT) in Model 1,;( (1) = 22.3,p < .001, correctly classifying 71.1% of
observations (see Table 15). Consumers exhibiting higher (vs. lower) PMV demonstrated
greater boycott behavior in the simulated shopping scenario, B = .67, SE = .15, Wald =
18.89,p < .001. However, a significant Hosmer-Lemeshow (HL) goodness-of-fit statistic,
;( (8) = 19.6,p < .05, suggested a less than ideal fitting model. A significant Hosmer-
Lemeshow (HL) value suggests that we reject the null hypotheses that there was no
67
difference between observed and predicted values. BI was also a significant predictor of
BOYCOTT (Model 2), X- (1) = 59.6,p < .001, also accounting for 71.1 % of correct
predictions (see Table 15). Unsurprisingly, BI was greater for boycotters than non-
boycotters, B = .87, SE = .14, Wald = 41.1,p < .001. The HL goodness-of-fit statistic was
non-significant, X- (8) = 7.93, p > .45, suggesting that boycott intention provided a better-
behaving model of boycott shopping behavior than perceived moral violation. A final
logistic regression model combined PMV and BI in the same model (Model 3), X- (2) =
59.79,p < .001, correctly accounting for 70.6% of observations. The HL goodness-of-fit
statistic was non-significant, X- (8) = 5.46,p > .71.
Table 15. Observed and Predicted Frequencies of Boycott Behavior (BOYCOTT)
Predicted
Observed No Boycott Boycott
% Con'ect
Modell
No Boycott
Boycott
Overaii % Correct
Model 2
No Boycott
Boycott
Overall % Correct
82
31
82
31
27
61
27
61
75.2%
66.3%
71.1%
75.2%
66.3%
71.1%
Model 3
No Boycott
Boycott
Overall ()'O C:orrcct
80 29 73.4\%
30 62 67.4%
'7i'l COl
1\.I.UfO
68
The path between PMV and BOYCOTT ceased to be significant with the
inclusion of B1 (illustrated in Figure 8). This pattern of significant paths suggested that B1
potentially mediated the relationship between PMV and B1 and was a more proximal
predictor of BOYCOTT than PMV (Baron & Kenny, 1986). To test for mediation, linear
regression was first used to test the relationship between the PMV and B1. The overall
model was significant, R2 = .34, F(1, 199) = 103.7,p < .001. Next, a mediation test with a
continuous mediator and a dichotomous outcome (Mackinnon & Dwyer, 1993; see
Appendix D Exhibit 1 and 2 for SPSS Code and complete computations for determining z
score) determined that B1 fully mediated the effect of PMV on BOYCOTT, Sobel z(197)
= 2.15,p < .05. Measuring boycott behavior in the real world may be difficult because it
is difficult to ascertain if a consumer decision at the point of purchase was due to boycott
or other factors. Thus, a potential proxy for actual boycott behavior in research may be to
measure boycott intention.
.26
(2.17)
PMV
HI
.17
( 1.00)
.81
(10.1)
Note: PMV == Perceived Moral Violation; BI == Boycott Intentions;
BOYCOTT::: Boycott Behavior (dichotomous). All path reported as
standardized coefficients; T-values are shown in parentheses; paths
denoted by solid lines are significant at p < .05 or better;
Figure 8. Boycott intention mediates perceived moral violation.
69
Model Testing
The hypothesized model was tested using path analysis in SEM. Generally, SEM
relies on maximum likelihood estimates (ML) that require continuous and normally
distributed variables; however, the indicator for boycott behavior (BOYCOTT) was
dichotomous. Dichotomous endogenous indicators violate the requirement for normal
distribution for SEM using ML (Byrne, 2010). Although the general analytic strategy is
to consider the dichotomous variable as an instance of a continuous variable or to use
biserial correlation matrices, Bayesian SEM provides a potential alternative (for a
complete discussion, see Byrne, 2010, pp. 148-160). The primary path analysis will use
ML estimation but Bayesian parameter estimates will be used to cross-validate the ML
parameter estimates.
The process model in Figure 9 was specified for path analysis in AMOS 7.0 using
both ML and Bayesian estimation procedures. In additional to specifying the
relationships outlined in the model, dummy-coded experimental manipulations were
included as exogenous variables directed to each of the endogenous indicators (excluding
HE and NABB, which are also exogenous variables in the model). Placing these
constraints on the model ensures statistical control for any differences attributable to the
manipulations. The model was trimmed after the initial fit assessment. All non-significant
paths between the manipulation dummy codes and endogenous variables that were not
part of the hypothetical model (as determined by the C.R. > 1.96) were eliminated from
the model. Model trimming is permissible when not used for ad hoc model development.
However, all paths in the hypothesized model were retained, even if the path was non-
significant. Figure 9 illustrates the path model but excludes the manipulation dummy-
coded variables for clarity. The ML parameter estimates are presented in Figure 6. A
complete list ofML and Bayesian parameter estimates is presented in Appendix F.
.52HE MEOther
.23
.13 (3.57)
.17 MESe1f ,,
,
(2.75) ,J)3\
(57) \
.07 ,,,
.38 (1.00) , .57,
PMV BI
I
I
I
I
,
I
.39 II,
.13 (6.41 ) I
.53I
(2.30) II (8.61 )
.51 " -.01BENE ,/ (-.25)
I
I
,
.04 I .27I
NABB COST BOYCOTT
Note: All path reported as standardized coefficients; T-values are shown in
parentheses, paths denoted by solid lines are significant at p < .05 or better;
italics = R2. HE = Humanitarian-Egalitarian Orientation; NABB = Negative
Attitude towards Big Businesses; MEOther = Other-condemning moral
emotions; MESelr Self-conscious moral emotions; BENE = Benefit evaluation;
COST = Cost evaluation; BI == Boycott Intentions; BOYCOTT= Boycott Behavior
(dichotomous).
Figure 9. Path diagram for hypothesized process model and estimated path coefficients.
The AMOS maximum likelihood (ML) results suggested that the hypothesized
model provided an acceptable fit to the data (X 2 = 73.7, df= 50, P < 0.05; RMSEA =
70
71
0.05; 90% CI = .02, .07; CFI = 0.96; SRMR = 0.083). The model accounted for 57% of
the variance in boycotting intention (BI) and 27% of the variance in boycott behavior
(BOYCOTT). Regarding the Bayesian model, the posterior predictive p-value = .41 for
the Bayesian model suggested the model fit was acceptable (Arbuckle, 2008). In
Bayesian models, a p-value of .50 suggests strong evidence of an exact model, whereas
models with a p-value of .05 indicate a poor fit. Models with p-values between these two
points suggest that the data fit portions of the model. The deviation from .50 in the
current hypothesized model (p = .41) may be attributable to the retention of two non-
significant paths in the hypothesized model. The two techniques demonstrated that model
fit was acceptable. More important, parameter estimate values were consistent across
both ML and Bayesian techniques.
All but three of the paths' in the hypothesized model were significant (ps < .05).
The path analysis indicated that greater levels of humanitarianism-egalitarianism
orientation (HE) and negative attitude toward big business (NABB) were positively related
to perceived moral violation (PMV) in boycott situations. People with greater
humanitarian-egalitarian orientation were more likely to feel greater perceived moral
violation in the boycott situation. Similarly, people with a greater negative attitude
toward big business were also likely to feel greater perceived moral violation when
confronted with the company's egregious actions. The results also suggested that people
were more likely to feel greater other-condemning moral emotions (e.g., anger, contempt,
disgust) and to perceive greater benefit to boycotting when they perceived greater moral
violation. PMV was positively related to other-condemning emotions (MEOther) and
72
benefit evaluation (BENE), which in turn was positively related to boycott intention (BI).
Consistent with the previous logistic regressions, BI increased the likelihood of boycott
behaviors (BOYCOTT). Thus, people who felt greater other-condemning moral emotions
and perceived greater benefit to boycotting were more likely to demonstrate greater
boycott intention and more likely to take boycott action. Self-conscious moral emotions
(MEself) were not predictive of boycott intention. In addition, perceived costs (COST),
although in the predicted direction, did not significantly contribute to BI.
The coefficient and significance of the indirect effects were estimated using
bootstrapping in order to compare the direct and indirect effects with multiple mediators.
Bootstrapping is necessary in order to generate SEs for indirect effects (Preacher &
Hayes, 2008). AMOS provides the bootstrap estimation of indirect effects and SEs as part
of the overall SEM output, so a separate analysis was not required (the default bootstrap
iteration was 200). Table 16 provides a summary of indirect effects in the hypothesized
model.
Table 16. Decomposition ofIndirect Effects from Path Analysis (Study 2)
Standardized Indirect Effect HE NABB PMV ME01he] BENE
MEathe] .10 .08
MEsclf .05 .04
BENE .07 .06
COST -.03 -.03
BI .06 .05 .31
BOYCOTT .03 .03 .20 .12 .21
All ps < .01
73
There were significant indirect effects on BOYCOTT from PMV (b = .20, SE =
.04,p < .01), MEOther (b = .12, SE = .04,p < .01), and BENE (b = .21, SE = .05,p < .01).
The mediated path through MEother (appraisal 7 emotions 7 intentions 7 behavior)
supported the proposition that boycotting may be conceptualized as a consumer coping
behavior.
Alternative Models
Study 2 data were used to model each of the alternative models shown in Figure
5. Humanitarian-egalitarian orientation (HE) and negative attitude toward big business
(NABB) were included in the models but are not shown in the diagram. These individual
differences were modeled as exogenous variables related to perceived moral violation
(PMV) in all models. Table 17 summarizes the fit statistic for each of the models.
Table 17. Alternative Models Fit Statistics (Study 2)
Model
,
df CFT RMSEA (90% Cl) SRMR AlC Fit Assessment)( p
Base 64.7 48 ns .98 .04 .01; .07 .07 176.7 Good
1 111.2 48 <.001 .92 .08 .06; .10 .09 223.2 Poor
2 196 52 <.001 .81 .12 .10; .14 .14 299.5 Poor
3 73.3 49 < .05 .97 .05 .02; .07 .07 183.3 Good
4 122 49 <.001 .90 .09 .07;.11 .09 231.8 Poor
5 56.7 47 ns .99 .03 .01; .06 .07 l70.7 Good
6 82.6 47 <.OOl .95 .06 .04; .OS .07 196.6 Moderate
Based on the fit criteria established a priori, two models were viable alternatives:
Model 3 and ModelS. Model 3 demonstrated good model fit, i(49) = 73.3,p < .05, CFl
= .97, RMSEA = ,OS, SRNIR = .07, AlC = 183.3. ModelS provided superior fit statistics
compared to the hypothesized model, i(47) = 56.7, ns, CFl = .99, RMSEA = .03, SRMR
74
= .07, AIC = 170.7. Model comparison between the hypothesized model (AIC = 176.7),
Model 3 (AIC = 183.3), and Model 5 (AIC = 170.7) suggested that Model 5 potentially
provides a superior performing competing model. In general, it appeared that models that
had the sequence moral emotions -7 cost-benefit evaluations (Model 3 and Model 5)
demonstrated a better fit to the data than the reverse, and suggested that an affect-driven
model provided a superior configuration for the variables in the model.
Path Analysis of Individual Process Components
Bayesian SEM was used to estimate path coefficients due to the large number of
indicators and the modest sample size. Even so, interpretation of path coefficients should
be interpreted with some caution. The path diagram in Figure 10 was specified in AMOS
7.0. According to the results, other-condemning moral emotions (MEOther) as well as
benefits (BENE) were the most significant predictors of boycott intention (BI).
Consistent with the descriptive data in Study 1, the data in Study 2 suggested that anger
and self-enhancement were the key predictors of boycott intention. In Study 2, boycotting
was an expression of moral anger and a way to publicly identify oneself as a moral
person. This perspective was reminiscent of Kozinets and Handelman's (1998)
description of boycotting as a morally transforming experience. Even though self-
enhancement was considered a benefit evaluation, an examination of the items
comprising the self-enhancement scale suggested that it may be better described as an
affective response than a cognitive response (i.e., boycott to relieve guilt, to feel better,
etc.). The data suggested that little attention was placed on self-conscious emotions (e.g.,
------------------
75
guilt, shame). Instead, fault was directly placed on the boycott target resulting in greater
other-condemning emotions. Other-condemning moral emotions (e.g. anger) appeared
cOiltribute directly to B1. In addition, neither counterarguments nor constrained
consumption contributed appreciably to B1. The variance accounted for by the cost
variables were very low (5% and 1%, respectively). These findings were consistent also
with the low rating for counterarguments and constrained consumption in Study 1.
.16
(2.79)
.23 (3.48)
.05(.1.00
--------------
.02 (.38)
--------------
t--=-'=--.>,..;;...;~;========~-----_..9_~ _U~)
.02 (.27)
--------------
-.03 (-.36
--------------
.02 (.36)
.04 (.56)
BI
.08 (1.30
MD .37 --------------
l---"--~:===S=E==.=52=: .39 (4.74 I
-.01 (-.09)CA .05 --------------~
~===C=C==~.O~lt -_·~~_~~·~_1)
Note: All path reported as standardized coefficients; T-values are shown in parentheses, paths
denoted by solid lines are significant at p < .05 or better; italics::: R2. HE ::: Humanitarian-
Egalitarian Orientation; NABB ::: Negative Attitude towards Big Businesses; PMV::: Perceived
Moral Violation; MEmher = Other-condemning moral emotions; MESelt=Self-conscious moral
emotions; BENE::: Benefit evaluation; COST = Cost evaluation; BI = Boycott Intentions;
BOYCOTT= Boycott Behavior (dichotomous). MD = Make a difference; SE = Self-enhancement;
CA = Counter-argument; CC = Constrained Consumption
Figure 10. Path diagram for individual process components (Study 2).
76
Summary
The data in Study 2 provided initial support for the conceptualization of
boycotting as a form of consumer coping behavior. The results supported the assertion
people with higher humanitarian-egalitarian orientation and a more pronounced negative
attitude toward big businesses were more likely to perceive greater perceived moral
violation in boycott situations. Perceived moral violation was found to predict moral
emotions and cost-benefit factors associated with boycott intention. The participant's
boycott intention was reflected in potential boycott behavior in an online shopping
situation. However, only other-condemning moral emotions and perceived benefit
provided significant paths from perceived moral violation to boycott intention. More
specifically, people tend to boycott when they are angry and to feel better about
themselves. The data did not support a causal link from either self-conscious moral
emotions or costs to boycott intention. Finally, comparisons of alternative models
suggested that an affect-driven model fit the data better; consequently, this provided
additional support that boycotting can be better conceptualized as a consumer coping
behavior rather than a cost-benefit evaluation.
Study 3
The purpose of Study 3 was to investigate the extent to which a firm could use
counter-messaging to reduce the likelihood that consumers would take boycott action as a
result of pro-boycott messaging, considering how simple it is for any organization to
issue pro-boycott messages in the media (e.g., Internet). Klein et al. (2004) proposed four
77
boycott counter-messaging tactics: (a) acknowledge actions and communicate positive
changes, (b) stand firm on course of action, (c) convey the negative repercussions of
consumer boycott action, and (d) focus on the positive aspects of the brand. However,
little is known about the effectiveness of each of these proposed tactics.
In Study 3, counter-message effectiveness was defined by a reduction in
consumers' boycott intention (BI). BI was the most proximal indicator of boycott
behavior in Study 2. For practitioners, minimizing BI should be the primary focus of
messaging efforts. From a practical messaging execution perspective, it may be difficult
to address all of the factors that influence BI simultaneously. Each of the proposed
counter-message tactics addresses specific processes that influence BI. The most sensible
approach is to attempt to manage the common antecedent in the model. In Study 2,
perceived moral violation (PMV) was a significant predictor of BI, mediated by multiple
emotional and cost-benefit factors. The hypothesized model proposes that the key to
deterring consumer boycott intention is to minimize PMV. Therefore, it is likely that the
Klein et al. (2004) tactic based on acknowledging actions and communicating positive
changes to reduce PMV should perform better than other tactics.
The analysis for Study 3 begins with a detailed manipulation check for the
counter-message press releases presented to participants. A set of diagnostic indices was
created to verify that the counter-message attributes identified by Klein et al. (2004) were
present and corresponded to a specific press release. Once the counter-message stimuli
were deemed to possess the attributes proposed by Klein et aI., the counter-message
condition was used as the independent variable in an ANOVA with BI as the dependent
78
variable. Finally, Study 3 replicated all of the path analyses in Study 2 to gain greater
insight into the hypothesized and alternative models.
Sample Characteristics
A professionally managed consumer panel (U.S. only) provided the participants
for Study 3. A total of 709 online surveys were usable after the data were screened. The
participants were screened in three ways for inclusion into the data set. First, proprietary
vendor attention and completion checks were included in the online survey. Only
participants that completed the survey and correctly answered the attention check were
included. Second, participants had to correctly answer a manipulation check question
regarding the target of the boycott at the end of the survey. Third, only participants that
have a positive attitude toward the boycott target (M = 6.20, SD = 1.06) and had recently
purchased branded products from the boycott target were included in the study. On
average, participants were frequent purchasers of chocolate products produced by the
boycott target (number of products purchased in the last month: M= 7.76, SD = 3.83).
The mean age (M = 34.6, SD = 11.7) was higher than prior samples. The greater range in
age potentially makes the finding of Study 3 more generalizable to a broader population.
Education and race were similar to prior studies (please refer to Table 7 in Study 1). For
example, participants were predominantly white Caucasians (n = 575, 81.1%) and college
educated (n = 471,66.4%).
79
Descriptive Statistics
Table 18 summarizes the descriptive statistics and provides the correlation matrix
for the variables included in the model. All variables were significantly correlated (all ps
< .05). The reliability of all scales was acceptable; the Cronbach a for all scales was>
.90 except for perceived cost (COST), Cronbach a = .74. Pearson correlations ranged
from .08 to .90. The strongest relationship existed between boycott intention (BI) and
perceived moral violation (PMV), other-condemning moral emotions (MEOther), self-
conscious emotions (MEse1f), and perceived benefit (BENE), r = .71, .82, .86, .90,
respectively. Both moral emotion groups (MEOther and MEself) were highly correlated (r =
.85). A strong relationship also existed between BENE and MEOther (r = .79) and MEse1f (r
= .83). This was not surprising considering that a key indicator in the BENE scale was a
desire to feel less guilt. In addition, PMV was highly correlated with BENE (r = .70),
MEOther (r = .72), and MEse1f(r = .63). All other variable relationships were below .50.
Conditions and Stimuli
The Global Exchange slave chocolate boycott was used as the boycott context for
Study 3. The experimental stimuli were a pro-boycott message and five counter-messages
presented as press releases. Participants were randomly assigned into one of six
conditions (summarized in Table 19). All press releases are presented in Appendix D. In
the control condition (Condition 1), participants read only the pro-boycott press release.
The press release called attention to the use of child slaves in the harvest of cacao beans
used by a leading chocolate company for American consumption.
Table 18. Descriptive Statistics (Study 3)
Intention Appraisal A10ral Emotions Evaluations Individual Dijferences
N=709 BI PMV MEother MEsc'f BENE COST HE NA BB
Bl 1
PMV .71 **
tv1E01!lcr .82** .72** 1
MEsdf .86** .63** .85**
BENE .90** .70** .79** .83**
COST -.18** -.16** -.14** -.08* -.16**
HE .41 ** .46** .45** .42** .43** -.09*
NABH .42 ** .46** .38** .37** .43** -.11 ** .33** ]
M 4.08 5.02 4.21 3.87 5.65 5.33 5.0 I 5.44
(SD) (1.97) (1.26) (1.82) (1.86) (2.92) (1.72) (0.80) (1.51 )
Items 6 4 3 4 7 6 10 1
a 0.96 0.92 0.90 0.94 0.96 0.74 0.92
Nofes:* p < .05; ** P < 0.01 level; BI = Boycott Intent; PMV = Perceived Moral Violation; MEOtlle]
= Other-condemning moral emotions (Anger, Contempt, Disgust); ME self = Self-consciolls Moral
Emotions (Guilt, Regret, Shame, Embarrassment); BENE = Benefit Evaluation (Make a differences,
Self-enhancement); COST = Cost Evaluation (Counter-argument, Constrained Consumption); HE =
Humanitarian-Egalitarian Orientation; NABll = Negative Attitude towards Big Businesses.
00
o
81
In the remaining conditions, participants were asked to read the pro-boycott press
release followed by one of five press releases issued by the targeted chocolate company,
which responded to the initial pro-boycott press release. In Condition 2, participants
received the original press release issued by the chocolate company. This condition was
used as a benchmark for the experimental counter-messages.
For Condition 3, the objective ofthe press release was to attempt to reduce the
consumer PMV by acknowledging its egregious practices and to communicate
ameliorating future actions. In Condition 4, the press release explained the situation and
declared that the company stood firm on its continued actions. The primary message
objective was to communicate to the consumers that the firm had heard their protests but
that the consumers' actions would not make a difference on company actions and
practices.
An alternate tactic was to convey the negative repercussions of consumer boycott
action (Condition 5). Based on Klein et al.'s (2004) model, the primary objective of the
press release was to provide convincing counterarguments (to increase perceived cost)
and removed the potential ability of consumers to feel good about themselves by
boycotting. By enumerating potentially unexpected negative consequences to the affected
party, the boycott action was framed to have unexpected high perceived costs with very
little self-enhancement benefits. The final condition (Condition 6) focused consumer
attention on the positive aspect of the firm and products and bypassed discussion
regarding any egregious action (Treatment 6). By drawing consumer attention to the
positive aspect ofthe brand, this counter-message sought to increase the perceived value
82
and reliance on the brand, and thereby increase the perception of constrained
consumption (Klein et aI., 2004).
AttributesTacticCondition
Table 19. Experimental Conditions and Counter-Message Tactics
Counter-Message
Objective
1 None Control n/a
2
3
Denial and
communicate
leadership
Acknowledge
actions and
communicate
amerliorating
changes
Original Press
Release
Reduce PMV by
acknowledging
actions and
communicating
positive changes
n/a
1) communicate changes in the
firm's practice,
2) provide an explanation for
practices,
3) communicate positive actions it
has taken to make the
situation better
4
5
Stand firm and
explain course
of action
Convey
negative
repercussions
of consumer
action
Decrease the ability
for consumers to
make a difference
Increase
counterarguments;
decrease self-
enhancement
1) communicate they have heard
consumer concen1S,
2) have chosen the bet course of
action,
3) will not change their
decision/practice?
1) convey negative repercussions
of the boycott,
2) credibly explain potential
negative outcome for large
scale boycott,
3) reduce the ability for the
consumers to feel good about
himself
6 Promote
positive aspect
of firm and
product
Increase constrained
consumption
1) discuss positive aspect of the
firm
2) dicuss positive aspect of
products
The experimental manipulation required participants to attend to the specific
details in each of the press releases. To verify that each press release correctly executed
83
the intended tactic, manipulation checks were gathered immediately after the participants
read the counter-message. The items presented in Table 20 were based on a 7-point
Likert scale (1 = Very ineffective, 4 = Moderately effective, 7 = Very effective). Each
manipulation check item corresponded to a specific counter-message tactic (Condition 3-
6). The manipulation check items were combined to create set of summary diagnostic
indices (CMI-4).
The manipulation check indices were analyzed separately to assess the success of
the manipulations in the experiment. Ideally, the index that corresponded to a specific
tactic should have the highest relative value compared to the other tactics. Table 21
provides the basic descriptive statistics of the created manipulation indices. Figure 11
provides a profile plot of each of the indices grouped by tactic. The control group was
excluded in the manipulation check analysis because this group did not receive a counter-
message. The profile for the original press release (Condition 2) was included for
comparisons. Visual inspection suggested that the counter-message tactics appeared to
have been executed successfully. The peak index values corresponded with the
appropriate tactics.
As shown in Figure 11, the index that corresponded to acknowledging actions and .
communicating changes in practices (CMl) appeared to be highest for the corresponding
counter-message tactic (Condition 3). A planned contrast suggested CMl was higher for
the acknowledge and communicate tactic relative to other tactics, M = 5.03, SD = 1.35,
t(587) = 2.86, p < .01.
84
Table 20. Summary of Manipulation Check Indices, Items, and Corresponding Tactics
Index Manipulation Check*
CMl Communicate changes in [Company's] practices
Communicate actions taken to make the situation better
Communicate that [Company] has heard the consumers'
concern
CM2 Provide an explanation for the [Company's] practices
Indicated that [Company] has chosen the best course of
action
Indicated that [Company] will not change its action
CM3 Convey negative repercussions of a boycott
Explained the potential negative outcomes of a boycott
Attempt to make [consumer] feel good [R]
CM4 Discuss the positive aspect of [Company] products
Discuss the positive aspect of [Company]
Tactic (Condition)
Acknowledge
action and
communicate
ameliorating
changes
(Condition 3)
Stand firm and
explain course of
action
(Condition 4)
Elaborate on
negative
reperCUSSIOns
(Condition 5)
Promote positive
aspect of firm and
product
(Condition 6)
* 7-point Likert Scale: Very ineffective-Moderately Effective-Very Effective
Similarly, the peak value for CM2 was for the tactic associated with reducing the
perceived efficacy of the consumer boycott against the company by standing firm on its
course of action and its practices (Condition 4). The CM2 index indicated the degree to
which the press release explained the firm's practices and stated that the firm would not
change it course of action. A planned contrast suggested that CM2 was highest for the
reduce efficacy tactic, M == 4.88, SD == 1.24, t(587) == 3.25,p < .001.
----------- ------------------
85
Table 21. Means for Counter-Message Manipulation Indices (CMl-4)
Index
CMl CM2 CM3 CM4
Acknowledge Reduce Negative Positive
Action Efficacy Repercussions Aspects
Counter-message n M SD M SD M SD M SD
Denial (Original) 123 4.84 1.37 4.52 1.34 4.10 1.45 4.59 1.56
Acknowledge Action 121 5.03 1.35 4.38 1.31 3.99 1.51 4.59 1.55
Reduce Efficacy 118 4.26 1.73 4.88 1.24 4.00 1.70 4.09 1.89
Negative Repmcussions 120 4.16 1.74 4.53 1.42 4.76 1.35 4.03 1.78
Positive Aspects 118 4.15 1.60 4.22 1.37 3.95 1.39 5.30 1.38
The CM3 index assessed the degree to which the press release conveyed the
negative repercussions of boycotting the firm and minimized the potential ability for
consumers to feel good about themselves by engaging in boycott actions. In Figure 11,
the index appeared highest for Condition 5, which corresponds with the negative
repercussions tactic; a planned contrast supported this finding, M = 4.76, SD = 1.39,
t(595) = 5.09,p < .001. There was clear correspondence between the CM4 index and the
final tactic based on the press release's positive portrayal of the brand and products. The
index for CM4 was highest for the tactic that emphasized the positive aspects of the firm,
M= 5.30, SD = 1.38, t(595) = 5.79,p < .001.
The profile for the original press release that was actually produced by the
chocolate company appeared to be nearly identical to the acknowledge action tactic (see
Figure 11) even though the execution of the press release was entirely different. The
chocolate company's tactic placated consumers by denying the pro-boycott allegations
and by providing contrary evidence. A set of planned contrasts that compared each index
value (CMl-4) between the original press release and the press release for the
86
acknowledgment tactic showed that the two tactics appeared indistinguishable from one
another, ts =.21 to 1.14, ns.
-CM1 ~CM2 ~CM3 DCM4
5.75
~
=
-; 5.25
;;;;
~
"'0
="""" 4.75
4.25
3.75
Denial
(Original)
Acknowledge
Action
Reduce
Efficacy
Negative
Repercussions
Positive
Aspects
Counter-Message Tactic
Figure 11. Profile of manipulation check indices for counter-message tactics.
Effectiveness of Counter-Messaging Tactics
The effectiveness of each of the counter-messages on influencing boycott
intention (BI) was examined after assessing whether the experimental manipulations
correctly activated the various counter-messages. An ANOVA test using planned
contrasts with BI as the dependent variable and six counter-message conditions as the
independent variable was significant, FeS, 703) = 3.42,p < .01, suggesting that there was
a difference between counter-message treatments and control conditions. When the
87
control condition was compared to counter-message conditions (see Table 22 for contrast
statistics), it appeared that any form of counter-message reduced BI, t(703) = 3.64,p <
.001. The simple act of responding to a company's egregious offense reduced BI. This is
consistent with the idea of "unthinking compliance" proposed by Cialdini (l998)-an
excuse, any excuse, placates people.
Table 22. Summary of Planned Contrasts for Boycott Intention (BI)
Planned Contrasts Value
Control vs. Treatments 3.64
vs. Original 0.77
vs. Acknowledge Action 0.90
vs. Reduce Efficacy 0.54
vs. Negative Repercussions 0.81
vs. Positive Aspects 0.61
Original vs. Acknowledge 0.13
Original/Acknowledge vs. Others 1.37
Note: *p < .05, ** p < .01, *** p < .001
SF
0.96
0.25
0.25
0.25
0.25
0.25
0.24
0.60
t (703)
3.78***
3.14**
3.62***
2.18*
3.26***
2.46*
0.52
2.27*
In Study 3, managing perceived moral violation (PMV) was apparently the most
direct way to reduce consumer boycott intention (BI). The correlation between PMV and
BI was r = .65,p < .001, and the plot for PMV mirrored that ofBI in Figure 12. This was
notable considering the difference in scale items for PMV and BI. PMV was a cognitive
appraisal and measured people's sense ofjustice and fairness, while the BI items
identified specific boycott activity. The tight coupling between these two constructs was
further examlned using path analysis.
6.00
5.50
'"
5.00
=~
Q,l
::;
4.50
4.00
3.50
,
"'
-BI
"'
""
....
-~ PMV
" ... - --+
88
Control Denial Acknowledge Reduce Negative
(Original) Actions Efficacy Repercussions
Counter-Message Tactics
Positive
Aspects
Figure 12. Effect of counter-messages on boycott intention (BI) and perceived moral
violation (PMV).
Model Testing
Figure 13 illustrates the path diagram that was specified in AMOS (similar to
Study 2). Appendix G provides the complete parameter estimates for the path diagram.
The standardized maximum likelihood (ML) estimates of the direct effects are presented
in Figure 13. The estimates of standardized indirect effects are presented in Table 23. The
coefficient and significance of the indirect effects were estimated using bootstrapping in
order to compare the direct and indirect effects. Bootstrapping was necessary in order to
generate SEs for total indirect effects (Preacher & Hayes, 2008). AMOS provides the
bootstrap estimation of indirect effects and SEs as part of the overall SEM output so a
separate analysis was not required (the default bootstrap iteration was 200). The results in
Figure 13 represent a test of the hypothesized model using the data from Study 3.
89
.02COST
.49
(17.8)
.51
BENE
.53HE MEOther
.41
.33 MESe1f
(10.3)
.10
.34 (4.49) .86PMV BI
.34
(l 0.4)
Note: All path reported as standardized coefficients; T-values are shown in
parentheses, paths denoted by solid lines are significant at p < .05 or better;
italics = R2. HE = Humanitarian-Egalitarian Orientation; PMV = Perceived
Moral Violation; NABB = Negative Attitude towards Big Businesses; MEOther=
Other-condemning moral emotions; MESelt= Self-conscious moral emotions;
BENE = Benefit evaluation; COST = Cost evaluation; BI = Boycott Intentions;
BOYCOTT= Boycott Behavior (dichotomous).
Figure 13. Path analysis of hypothesized process model using Study 3 data.
The model fit to the data was very good, i = 43.0, df= 35,p < .001, CFI = 1.00,
RMSEA = .02 (90% CI = .01; .03), SRMR = .02. All paths were significant in the path
diagram. The model accounted for 86% of the variance in boycotting intention (BI). The
significant indirect effect of PMV on BI (.54) was greater than the direct effect of PMV
on BI (.10), suggesting that much of the PMV was expressed through other-condemning
90
moral emotions (MEOther), self-conscious moral emotions (MEse1f), perceived benefits
(BENE), and perceived costs (COST).
Table 23. Decomposition ofIndirect Effects from Path Analysis (Study 3)
HE NA nB PMV
PMV
MEnthol .22 .22
MEsd! .19 .19
BENE .20 .20
COST -.06 -.06
BI .34 .25 .54
Allp <.01
Three paths in the hypothesized model that were not significant in Study 2 were
significant in Study 3. MEself appeared to be positively related to BI. This relationship
suggests that the greater the degree of guilt, regret, shame, and embarrassment felt by the
consumer as a result of the boycott situation, the greater the boycott intention. COST had
the predicted, albeit small, inverse relationship with BI. Consistent with Klein et al.
(2004), the greater the perceived cost due to either constrained consumption or
counterarguments, the less likely consumers are willing to boycott. Finally, the path
between PMV and BI was significant. This suggests that a direct relationship exists
between PMV and BI that is not related to emotions and cost-benefit factors. There may
exist heuristics that link the appraisal to intention that bypass the intervening processes,
or another variable or process not accounted for in this model that should be examined in
future research.
91
Consistent with the path analysis in Study 2, people with greater humanitarian-
egalitarian orientation (HE) and negative attitude toward big businesses (NAss) were
more likely to report greater PMV. PMV was a significant predictor of greater MEOther,
MEself, and BENE. In addition, HE, NAss, and PMV accounted for a good portion of the
variance of MEOther (53%), MEse1r(41%), and BENE (51%). Perceived cost (COST)
continued to explain little variance in the overall model. Counterargument and
constrained consumption had contributed little to boycott intention (BI) across three
studies. The relationship between PMV and COST was weak (b = -.17, SE = .03, p <
.001), and the relationship between COST on BI was weaker (b = -.04, SE = .03,p <
.01). Only 2% of the variance of COST was explained by PMV, HE and NABS. This has
been a consistent theme from Study I and Study 2. It is likely that in the context of easily
substitutable products, COST may not be as important a variable as perceived boycott
benefits (BENE). In the boycott situations presented in the experiment, it is possible that
consumers paid attention primarily to benefits such as making a difference and self-
enhancement benefit ("What's in it for me?" and "What's in it for the affected party?").
Counterarguments and constrained consumption associated with the boycott may be more
of an afterthought elicited by the survey itself. Additional research on the relationship of
COST and BI is needed to clarify if it is related to the boycott context or if it is related to
a bias of people being more attentive to personal gains than costs. The findings
pertaining to costs in this dissertation differed from previous research (i.e. Klein et aI.,
2004), and more research is needed to clarify the role of perceived cost in consumer
boycott situations.
92
Alternative Models
A series of alternative or competing models were considered in addition to the
hypothesized model (shown in Figure 14). The models are useful because they allow for
competing models that vary on the directionality of variables to be tested.
Modell Modell
Model 3 Modcl4
Model 5 Modcl6
Note: PMV::: Perceived Moral Violation; MEmher= Other-condemning moral
emotions; MESelr Self-conscious moral emotions; BENE::: Benefit evaluation; COST
= Cost evaluation; BI = Boycott Intentions; HE and NA not shown.
Figure 14. Structural configuration of alternative models (Study 3)
93
Humanitarian-egalitarian orientation (HE) and negative attitude toward big
business (NABB) were included in the models but are not shown in the diagram for
purposes of parsimony. These individual differences were modeled as exogenous
variables related to perceived moral violation (PMV). The diagrams in Figure 14 are
identical to the alternative path models tested in Study 2 (Figure 5) with the exception
that boycott behavior was not included in the diagrams for Study 3. Unlike Study 2, the
primary dependent variable in Study 3 is boycott intention (Bl). Table 24 summarizes the
fit statistic for each of the competing models.
Table 24. Alternative Models Fit Statistics (Study 3)
Model X df p CFI RMSEA (90% CI) SRMR
Base 43.0 35 < .001 1.00 0.02 .01; .03 0.02
1 251.0 35 < .001 0.95 0.09 .08;.1 0 0.12
2 1688 41 < .001 0.63 0.24 .23; .25 0.16
3 304 37 < .001 0.94 0.10 .09; .11 0.04
4 410 37 < .001 0.92 0.12 .12; .13 0.04
5 106 36 < .001 0.98 0.05 .04; .06 0.03
6 250 36 < .001 0.95 0.08 .07; .1 0 0.06
AlC
129.0
337.0
1761.8
385.8
491.7
190.2
323.6
Fit Assessment
Good
Moderate
Poor
Moderate
Poor
Good
Moderate
Based on the fit criteria established a priori, only Model 5 provided a viable
alternate model to the baseline model. Model 5 demonstrated very good fit to the data,
;((36) = 106.0,p < .001, CFl = .98, RMSEA = .05, SRMR = .03, AlC 190.2. Model 5
was structurally consistent with the consumer coping models (i.e., Duhachek, 2005).
Specifically, the primary feature of Model 5 is the direct relationship from moral
emotions to boycott intention, perceived benefit, and perceived cost. This relationship
suggests perceived benefits (BENE) may be part of a coping response to certain moral
94
emotions. For example, the self-enhancement items, such as "1 wouldfeel guilty" and "1
wouldfeel better about myselfif1 boycott," are responses that are more emotional than
economic. The alternative model results in Study 3 mirror the findings in Study 2, where
Model 5 provided a superior fit to the data. Reproducing the same alternative model
across two studies and two samples provides good corroborative support for a consumer
coping model.
Discrete Emotions and Individual Cost-Benefit Factors
Each of the moral emotions and cost-benefit factors were specified in the path
model as independent variables (see Figure 15). The model fit was very good, i = 191.5,
df= 94,p < .001, CFI = .99, RMSEA = .04, SRMR = .05. The model accounted for 87%
of the variance for boycott intention (BI). Appendix G provides the complete list of
parameter estimate. The indirect effect of perceived moral violation (PMV) on boycott
intention (BI) was significant through all moral emotions except for disgust and shame.
In both Study 2 and 3, self-enhancement (SE) provided the strongest link between PMV
and BI. Making a difference was a significant predictor of boycott intention in Study 3,
whereas it was not in Study 2. Constrained consumption was again non-significant as in
Study 2, and counterarguments did not appear to be related to boycott intention to a great
extent.
The importance of different moral emotions and cost-benefit factors appeared to
differ depending on the type and context of the boycott. There was a richer set of moral
emotions and cost-benefit factors that contributed to overall boycott intention in Study 3.
---------_.__ .-
95
While other-condemning moral emotions (MEOther), especially anger, played a greater
role in boycotts in Study 2, the chocolate boycott in Study 3 appeared to be more reliant
on self-conscious emotions such as guilt, regret, and embarrassment.
.60 (19.7) Anger .36 1 .08 (2.67)HE
.61 (19.3) .05(2.21)
Contempt .36
.68(24.8)
Disgust .46
.08 (.63)
--------------
.60 (19.91 Guilt .36 .13 (4.50)
.33
.56 (18.1) .09 (3.83)(10.3) Regret .32
.55[17.6) .1 Shame .31 .02 (.77)--------------
.59 (21.9)
Embarrassment .34
1
.07 (2.24
.34 .34 .08 (4.09)
.
87
1PMV BI
~.V .64 (22.4) .I .11 (4.19(10.4) MD .41~..69 (25.8) SE .48 1 .43 (13.0-I
-.05 (-4.191 CA .02 -.05 (-3.79
-.01 (-.26JNABB i -.01 (-.04) ~
------------ CC .00 ~ -- - -- -- -- --- --,
Note: All path reported as standardized coefficients; T-values are shown in parentheses,
paths denoted by solid lines are significant at p < .05 or better; italics = R2. HE =
Humanitarian-Egalitarian Orientation; NAss = Negative Attitude towards Big Businesses;
PMV = Percevied Moral Violation; MEOther = Other-condemning moral emotions; MESe1r
Self-conscious moral emotions; BENE =Benefit evaluation; COST =Cost evaluation; BI =
Boycott Intentions; BOYCOTT= Boycott Behavior (dichotomous). MD = Make a difference;
SE =Self-enhancement; CA = Counter-argument; CC =Constrained Consumption
Figure 15. Path analysis of individual emotional and cost-benefit factors (Study 3).
96
Summary
The results of Study 3 suggested that boycott intention could best be minimized
by managing a consumer's perceived moral violation; however, it appeared that any form
of counter-messaging (with the exception of, perhaps, the reduced efficacy tactic)
reduced boycott intention. Further, the performance of simply denying allegations of
wrongdoing (original chocolate company response) may be as effective as admitting
wrongdoing and taking corrective action.
Path analysis in Study 3 replicated and extended the findings in Study 2. Whereas
other-condemning moral emotions (e.g., anger and contempt) were a stronger predictor of
boycott intention in Study 2, the contributions of self-conscious moral emotions (e.g.,
guilt and regret) played a greater role in the Study 3. Analysis of individual emotional
and cost-benefit factors suggested that moral emotions and benefits were key predictors
of boycott intention. Anger, guilt, make-a-difference, and self-enhancement were the
components that contributed the most to boycott intention. Both counterarguments and
constrained consumption demonstrated weak relationships with perceived moral violation
and boycott intention, consistent with findings in Study 1 and 2. Further investigation is
needed to better understand the role of cost in the overall model.
Similar to the finding in Study 2, comparisons of alternative models suggested
that an affect-driven configuration provided better fitting models. The replication of the
directionality of the relationships based on the alternative model configuration with
independent samples in Study 2 and 3 provided support that the boycotting process may
more resemble consumer coping behavior than a cost-benefit assessment.
--------------------- - -~-
97
CHAPTER V
DISCUSSION
The phenomenon of consumer boycotts is multifaceted. It is therefore not
surprising that the study of consumer boycotts is influenced by research in areas such as
economics (Chavis & Leslie, 2009; John & Klein, 2003), brand transgressions (Aaker,
Fournier, & Brasel, 2004), corporate social responsibility (Brown & Dacin, 1997; Sen &
Bhattacharya, 2001), ethical consumption (Cooper-Martin & Holbrook, 1993), prosocial
behavior (Bagozzi & Moore, 1994; Sen, et aI., 2001), and consumer activism and
retaliation (Gregoire & Fisher, 2006; Gregoire & Fisher, 2007; Kozinets & Handelman,
2004). The relatively small body of research on the topic primarily defines consumer
boycotts with a traditional economic framework. In this view, boycotts are collective
social actions predicated on individual cost-benefit analysis: consumers will withhold
purchase if they perceive that eventual boycott outcome benefits outweigh immediate
short-term sacrifices. This view may be sufficient to describe the analytical processes
involved in more traditional boycotts; however, modem boycotts differ because the
primary boycott motives and objectives are emotional rather than economic.
The current research contributes to boycott research by presenting an augmented
consumer boycott model. The hypothesized model extends the older cost-benefit model
by incorporating a complementary emotional process. In this new boycott model, the two
98
processes are thought to simultaneously contribute to the formation of consumer boycott
intention. Specifically, the first process is the cost-benefit analysis. Consumers base their
decision to boycott on individual evaluations of perceived costs and benefits. The second
process is a coping mechanism based on moral emotions (Haidt, 2003; Tangney,
Stuewig, & Mashek, 2007). In this process, consumer boycott intention is a response to
the emotional distress generated by a boycott situation. The resultant two-process model
is reminiscent of other dual-process models that have an emotional and a rational
component such as the affective-cognitive model of consumer decision making (Shiv &
Fedorikhin, 1999), the Elaboration Likelihood Model of Persuasion (Cacioppo & Petty,
1984) or the two-system model of intuition and reasoning (Kahneman, 2003).
The new model expands on the ideas of Klein et al. (2004) and Sen et al. (2001).
In their research, consumer boycotts were framed as prosocial behaviors, but little
attention was paid to the crucial role that affect and coping processes play in prosocial
behaviors. Therefore, the model considered in the current dissertation explicitly
incorporates affective factors hypothesized to influence boycott decisions. These factors
include perceived moral violation (an appraisal), moral emotions, and coping (boycott
intention). When consumer boycott is conceptualized as consumer coping, boycott
intention reflects a consumer's emotional response to the social injustice arising from a
preferred brand's egregious actions that harms others.
Three studies in the current dissertation provide support for the hypothesized
model. In an online survey of active boycott participants, Study 1 suggests that boycotts
are emotionally complex situations involving other-condemning moral emotions (anger,
99
contempt, and disgust), self-conscious moral emotions (guilt, regret, embarrassment, and
shame), and prosocial emotions (hope, empathy, and sympathy). In addition, emotions
such as betrayal, trust, suspicion, and hatred are also relevant in boycott situations. Study
1 contributes to the boycott literature by identifying a list of emotions that have been
overlooked in the boycott literature.
Study 2 links together two constructs supported in previous research: perception
of moral violation and boycott intention (Klein, et al., 2004) and boycott intention and
boycott behavior (Sen, et al., 2001). The results indicate that boycott intention mediates
the direct effect of perceived moral violation on boycott behavior. The primary emphasis
of Study 3 was to examine the applied implications of the hypothesized model. The
effectiveness of a variety of counter-messaging tactics previously proposed by Klein et
al. (2004) was tested. Study 3 confirms that managing perceived moral violation is the
best way to manage boycott intention. Interestingly, any form of counter-message
appears to reduce perceived moral violation, but the tactic of communicating positive
change and acknowledgement, either denying wrongdoing or admitting wrongdoing, is
most effective in reducing boycott intention.
A series of path analyses using the data from Study 2 and Study 3 supported the
hypothesized two-process boycott model. The findings of the path analyses are consistent
across both experiments. The use of path analysis offers three distinct advantages. First, it
permits the simultaneous calculation of parameter estimates of direct and indirect effects
hypothesized in the model. Second, it provides insights into the contribution made by
individual predictors (discrete emotions and cost-benefit factors) to boycott intention.
100
For example, other-condemning moral emotions and perceived benefits were the most
significant predictors of boycott intention in Study 2, whereas self-conscious moral
emotions and perceived benefits were most predictive in Study 3. Overall, anger and guilt
specifically appeared to be strong indicators of boycott intention. Finally, path analysis
allows for the test of competing models that can then be compared to the hypothesized
model.
A summary of key findings along with the theoretical and applied implication is
presented in the remainder of this chapter for perceived moral violation, moral emotions,
and the general indication that an affect-driven model is the proper configuration for the
hypothesized model. Chapter V concludes with a discussion of the limitations of this
research along with opportunities for future research.
Perceived Moral Violation
Perceived moral violation is a crucial determinant in consumer boycotts. The
pattern of moral emotions, cost-benefit factors, boycott intention, and boycott behavior
across three studies indicates that the overall consumer boycott model is contingent on a
consumer's heightened sense of moral violation. This finding is consistent with prior
research (i.e., John & Klein, 2003; Klein, et aI., 2004; Sen, et aI., 2001) and has even
been recognized in the cost-benefit framework. For example, although never tested,
Tyran and Engelmann (2005) have noted that "fairness considerations may be of
considerable importance" (Tyran & Engelmann, 2005, p. 213).
101
In appraisal theory, the perception of social justice or fairness has been
conceptualized as a dimension referred to as legitimacy or normative and moral
comparability (Johnson & Stewart, 2004; Mauro, et ai., 1992). Haidt (2001) proposed
that moral judgments are "caused by quick moral intuitions and is followed (when
needed) by slow, ex post facto moral reasoning" (p. 817). Moral intuition isa form of
cognition that involves less elaboration than reasoning. Thus, perceptions and intuitions,
followed by emotions and reasoning, contribute to the formation of boycott intentions.
Perceived moral violation and boycott intention appear to be tightly coupled. In
Study 2, perceived moral violation was positively related to boycott intention and boycott
behavior. Moreover, boycott intention was found to fully mediate the effect ofperceived
moral violation on boycott behavior, thus indicating that intention is a more proximal
indicator of behavior than perceived moral violation. The mediation results also confirm
the strong relationship between perceived moral violations and boycott intention. This
effect was replicated in Study 3, where a reduction in perceived moral violation was
found to reduce boycott intention in a nearly monotonic fashion.
Path analyses in Study 2 and Study 3 provide additional insights into the
relationship between perceived moral violations and boycott intention. The indirect effect
of perceived moral violation through both moral emotions and cost-benefit variables is
greater than the direct effect of perceived moral violation on boycott intention. This
finding suggests that intervening emotional and cost-benefit processes together help
account for the considerable amount of variance in boycott intention. More importantly,
102
the findings support the key proposition that an emotional process contributes to the
overall formation of boycott intention independently of the cost-benefit analysis.
In addition, a contribution to the boycott literature was the identification of two
individual difference variables found to influence intention to boycott. Humanitarian-
egalitarian orientation and negative attitude toward big businesses were both significantly
related to perceived moral violation in boycott situations. In Study 2, a significant
difference was found between boycotters and non-boycotters in humanitarian-egalitarian
orientation. Those scoring higher in humanitarian-egalitarian orientation were more likely
to be boycotters than non-boycotters. Further, path analyses in Study 2 and Study 3
provide support that both individual differences had significant indirect effects on boycott
intention. Respondents who held greater humanitarian-egalitarian orientation and
negative attitude toward big businesses perceived a greater degree of moral violation,
formed stronger boycott intention, and had greater tendency to take boycott action.
The managerial implications of the key findings are straightforward. The
management of perceived moral violation is the most important task for corporate public
relations and boycott organizers. For public relations practitioners, timely and appropriate
response to any pro-boycott message is necessary. Although it appears that any counter-
message tactic will do, counter-messages aimed at reducing moral violation through
acknowledgement (or denial) of the egregious act and communicating ameliorating
actions are more effective in reducing consumer boycott intention. Although the specific
tactic to reduce perceived moral violation might differ, the potentially detrimental effect
of trying to deceive the consumer (for example, denial or lying) may be less beneficial in
103
the long term. Instead, simply acknowledging wrongdoing and making reparations may
be the simplest way to proceed. In addition, based on the qualitative responses in Study 1,
simply conceding and making amends to boycotters demand to resolve the boycott may
not be sufficient to end the consumer's intention to boycott. The data suggest that it is not
uncommon for boycotters to continue to boycott even after boycott issues are resolved.
The implication of this is that consumers must be wooed back as new customers.
For boycott organizers, perceived moral violation is predicated on the belief that a
corporation has abused its power. Boycott activists are more likely to generate the moral
emotions required to engage a coping response leading to a boycott intention by issuing
messages that contrast the company's intent, size, and power with the disenfranchised
affected parties that are unable to protect themselves.
Moral Emotions
The current research suggests that it is imprudent to exclude the consideration of
emotions in models of boycott intentions. Emotions are distinct variables that need to be
measured independently. It is unfortunate that earlier research did not incorporate
emotions since moral emotions appear to be a consistent predictor of boycott intention. In
Study 1, other-condemning moral emotions (as well as prosocial emotions) differentiated
between symbolic and instrumental boycotts. Emotion items were rated significantly
higher in symbolic boycotts than in instrumental boycotts. This signifies the greater
emotional reactivity in boycotts that are motivated by a need for expression, punishment,
and clean hands. The findings in Study 2 and Study 3 are consistent with Haidt's (2003)
104
assertion that anger and guilt are prototypical other-condemning moral emotions that
elicit greater prosocial action tendency in support of another. The path analyses also
replicate Montada and Schneider's (1989) findings that anger and guilt are significant
predictors of prosocial action as the consequence of perceived injustice. In Study 2, anger
is the only moral emotion significantly related to boycott intention. The relationship
between self-conscious moral emotions and boycott intention did not achieve
significance. This is not surprising because all of the stimuli in Study 2 focused blame on
the target brand, so other-condemning moral emotions had greater influence on boycott
intention than self-conscious emotions. In Study 3, both other-condemning emotions
(anger and contempt) and self-conscious moral emotions (guilt and regret) were
significantly related to boycott intention as predicted in the hypothesized model. Overall,
the general results of the studies presented here support Haidt's (2003) assertion that
moral emotions lead to moral action. Boycott intention is influenced by the emotional
distress associated with social injustice and moral violation.
An Affect-Driven Model
The primary finding from the current dissertation appears to be that boycotting
strongly resembles a consumer coping process. A two-process model that incorporates
moral emotions and cost-benefit factors is a good starting point for model development.
This configuration provides the ability for other-condemning and self-conscious moral
emotions and perceived benefit and cost to directly influence boycott intention. The
hypothesized model was partially supported in Study 2 and fully supported in Study 3.
105
However, the path analyses of alternative models that test the directionality and sequence
of the variables in the model strongly suggest a distinct hierarchical ordering. In both
Study 2 and Study 3, the affect-driven models appear to perform better relative to the
other alternative models. Specifically, ModelS (see Figure 13) provides the best
alternative model to the hypothesized model in both Study 2 and Study 3. In ModelS,
moral emotions precede cost-benefit factors and moral emotions are directly related to
boycott intention. In this model, cost-benefit factors are conceptually part of the
consumer coping process described by Duhachek (2005) as rational thinking where a
consumer deliberately "attempts to prevent subjective emotions from directing behavior"
(p.44).
Model 5
Note: PMV = Perceived Moral Violation;
MEOther= Other-condemning moral
emotions; MESelr Self-conscious moral
emotions; BENE = Benefit evaluation;
COST = Cost evaluation; Bl = Boycott
Intentions.
Figure 16. Best fitting alternative model (ModelS)
---------c--- -~---------- --
106
Although both Model 5 and the hypothesized models are coping process models,
Model 5 provides superior fit in Study 2 (which includes a behavioral measure for
boycotting) and is more structurally consistent with the configuration of other affective-
cognitive models. For example, the elements included in the affect-driven model are
consistent with those highlighted in Kahneman's (2003) two-system model in which a
rapid affective system (System 1) is engaged initially and this is then followed by a
slower more deliberate process of reasoning (System 2) when the situation calls for
greater elaboration. Both Haidt's (2001) moral emotion theory and Duhachek's (2005)
consumer coping behavior are derivations of the two-system model. In Haidt's model of
moral judgment, moral intuition (moral emotions) is fast and effortless, and is followed
by a slower moral reasoning process when necessary. In Duhachek's emotional-cognitive
model of coping, emotions are thought to lead to more rational thinking. The affect-
driven boycott model tested in this dissertation incorporates similar features of these three
models if a cost-benefit analysis is considered to be equivalent to a slow and deliberate
reasoning process. Thus, the initial appraisal of perceived moral violation and subsequent
emotions form the rapid System 1 response, and the cost-benefit analysis characterizes
the slower System 2 (moral reasoning, rational thinking/cognitive perceptions) response.
In conclusion, affective components such as cognitive appraisals and emotions
appear to lead the way in boycott situations. It is likely that consumers participate in
boycotts because they are angry at a company's egregious act or feel guilty about
consuming products that contribute to harming others. Because existing cost-benefit
models do not explicitly account for the possibility that consumers may boycott for
--------------
107
reasons other than economic benefits and costs, the current research proposes new
boycott models based on both emotions and cost-benefit factors to close this gap.
Limitations and Future Research
Several shortcomings in the current research provide opportunities for future
research. Consumer boycott research can greatly benefit from continued boycott research
in four areas. First, better measures of perceived costs and benefits are needed to build a
more comprehensive consumer boycott model. Second, additional research is needed to
examine how moral emotions in boycott situations may vary across product category and
boycott types. Third, better boycott behavior measures are needed for marketing research
and application. Fourth, continued research is needed to better understand the sequence
and timing of messages and counter-messages and how they play out, since the
marketplace has become a social forum for consumer activism.
It was difficult to clearly establish a relationship between perceived cost
(counterarguments and constrained consumption) and boycott intention. In Study 1,
constrained consumption scored low relative to the midpoint of the scale and
counterarguments scored near the midpoint, suggesting that these two evaluations
contributed little to consumers' motivation to boycott. Further, the relationship between
the perceived costs variable and boycott intention was non-significant in Study 2 and was
very weak in Study 3. Another indication of the weak relationship was that very little of
the variance for perceived costs was accounted for in the overall model in either Study 2
or Study 3 (less than 5% in all instances). One interpretation may be that perceived costs
108
have little role in the overall model and that consumers attend only to perceived benefits.
Such an interpretation would be contrary to earlier models where costs were accounted
for in economic terms. Another possibility is that the perceived costs variable was
misspecified and may function differently than what was tested in the hypothesized
model. Klein et al.'s (2004) measure for perceived economic benefits was more related to
emotional response than economic value, so modeling perceived costs as the opposite of
benefits may not be entirely appropriate. Additional research is needed to examine how
the variable for perceived costs is related to other variables in the model, if such
relationships exist. Quite simply, a better set of measures for economic costs and benefits
are needed.
More research is needed on discrete moral emotions and moral emotion groups.
Moral emotions were not individually manipulated in any of the studies. A follow-up
study in which a few discrete moral emotions (for example, anger and guilt) are directly
manipulated is needed to provide a better understanding of the role played by moral
emotions in boycotting intention and behaviors. In addition, the individual contributions
made by other-condemning and self-conscious moral emotions to boycott intention
appeared to differ vary from one boycott situation to another in the current research.
Future research is needed to better understand how product category, boycott types, and
different situations may affect each of these emotion groups.
Very few behavioral measures for boycotting exist because the behavior is
difficult to directly observe. It is difficult to conclusively tie sales figures (Chavis &
Leslie, 2009; Klein, et aI., 2004) to specific boycott activities or advocacy messages due
109
to extraneous factors that exist in real-world data. Behavioral intention measures have
demonstrated some success in predicting consumer boycott intention but only in
simulated environments (Study 2; Sen, et aI., 2001). In the end, boycott decisions may be
made at the point of sale. Potentially, new technology that includes both measures of
boycott intention and the ability to track consumer purchase behavior (i.e., retail loyalty
programs and scanner-based tracking records) over extensive periods of time may
provide a more conclusive measure for actual boycott behavior.
An important aspect of messaging and counter-messaging was not fully explored
in this dissertation. The order of messages and follow-up messages provide an interesting
avenue of research. In the experimental studies, only pro-boycott messages and corporate
counter-messages were provided. If a company knows in advance that it cannot avoid
certain actions, it is possible for marketing communications and public relations to
preemptively issue counter-messages in advance of pro-boycott messages (Klein, et aI.,
2004). A company may choose to take the lead in order to inoculate its audience against
increased levels of perceived moral violation. Another possibility is that pro-boycott
supporters may issue a public rebuttal to the corporate counter-message. This could be
especially damaging if the corporate counter-message denied allegations or used
deceptive tactics and the rebuttal called attention to the company's less than truthful
intentions. A key question is whether preemptive and follow-up messages can increase or
reduce perceived moral violation.
In addition to understanding how the perceived moral violation, moral emotions,
cost-benefit factors, and boycott intention relate in a simple message/counter-message
110
scenario, further research regarding the ordering of messages, counter-messages, and
counter-counter-messages, etc., may provide additional insight into the workings of the
marketplace as a social forum (Kozinets & Handelman, 2004). The timing of the
response may playa role in affecting perceptions of moral violation. For example, the
one-year delay Toyota made in acknowledging its defects may have significantly
damaged their brand image (Healey, 2010; Philips, 2010). Recalls are a regular
occurrence for major automobile companies. Toyota's communication delay can be
contrasted with timely and public communication in the same types of recall situations
faced by other manufacturers. Immediate and honest communication along with
ameliorating actions generally diffuses consumer anger with little media fanfare. Thus,
message sequence and timing, in addition to counter-message tactic, would be a natural
next step in the investigation of the application of the hypothesized model.
Postscript
Perhaps it is naIve and overly optimistic to assert that research into consumer
boycotts, perceived moral violations, and moral emotions can provide insight into the
human condition. However, the results of the current research suggest that modem
consumers care about the welfare of others rather than just their own self-centered
economic interests. When confronted with a social injustice, it appears that consumers are
moved by personal convictions and emotions to take action to correct the injustice. In our
materialistic consumer culture, there may be the seed of doing good even in the
consumption of basic goods.
111
APPENDIX A
SUMMARY OF PRIOR BOYCOTT RESEARCH
t E fh d R fItt RfP' BT bi A 1 Sa e - urnrnary 0 nor oyco esearc an e erences 0 rno IOns
Level of Perspective! Emotions
Article Year Analysis Orientation Method Theory Referenced
Mahoney (1976) 1976 cn Survey
Miller & Sturdivant 1977 Individual, cn Survey, Panel CSR
(1977) Firm (Natural
Experiment)
Witkowski 1989 Individual cn Historical Descriptive
Kozinets & 1998 Individual cn Ethnography Interpretative Guilt, "Morality"
Handelman (1998)
Sen, Zeynep & l2001 Individual cn Experimental Cost-Benefit,
Morwitz (2001) Social
Dilemma
John & Klein (2003) l2003 Individual, CB Dynamic Cost-Benefit Anger, Guilt,
Group Economic Outrage
Model
Klein, Smith, & 2004 Individual, cn Survey Cost-BenefIt GuiJt*
Andrew (l2004) Group (Natural
Experiment)
Farah (Working) Individual CB Survey [Theory of
IPlanned
Behavior
Smith (1990) 1990 Consumer Case Studies CSR 'Anger, Guilt~
Policy & Moral Outrage
Activism
IFriedman (1985, 1999 Consumer Historical, VJnger, Moral
1991, 1995, 1999) Policy & Secondary, Outrage
A ,..-t-: ... T;"rt... (;:111"-"0'"
1 """\'1 V lulll ~-'Ul V,,",)
Innes (l2006) ~OO6 Firm F,conomic Economic Asymmetric
Modeling Information
------------ ._-~-
112
Level of Perspective! Emotions
Article Year Analysis Orientation Method Theory Referenced
Tyran & Engelmann 2005 Economics IExperimental
(2005)
Garrett (1987) 1987 Marketing Survey,
Management Secondary
* Measured
APPENDIXB
STUDY 1 SURVEY INSTRUMENT
113
114
Default Question Block
Background: We are interested in understanding why people participate in boycotts from a theoretical
perspective. This is an exploratory study conducted by Johnny Chen, a graduate student from the University of
Oregon, as part of greater research that focuses on prasadal advocacy and welfare. II is the observation of this
research leam that 100 itttle acadernic research has looked into how to consumer research can benefit the
consumer and social actlvisls, The goal of Ihis research is to beller understand how activist groups can more
effectively promote prosocial causes and boycotts as weH as understand how to "counter" counter-messaging
from companies that oppose the cause.
Instructions: For the following survey, we would greatly appreciate your input on one specific boycott that you
are most active in. Your input in this exploratory research wiff be used 10 creale a more general purpose survey
thaI can be administered 10 boycoll partiCIpants in the fulure. The total time commitment fm the survey is 20-30
minutes.
To maintain anonymity, we will not ask for your name in association with the responses. All information will be
coded an analyzed without any reference to you~ This information will NOT be shared with commercial or
governmental agencies. This survey is pure basic research. Your input and assistance will benefit science,
If you feel uncomfortable answering any of the questions, please leave it blank or on the default selting. You are
in no way obligated to complete Ihe survey and you may quit at any time. Any data entered up to the time Ihat
yOll quil will be removed
if you have ANY questions/issues regarding this surveyor simply wish to confimllhallhis survey is no! a scam of
any form, please conteel Johnny Chen at the University of Oregon via email (jchen3@uoregon.edu) or call
directly af +1 541 728-0930. I would be happy to address any questions or issues personally,
Boycott Participation
Please prOVide the name of a nation or company that you are currentty boycotting If you had previously
participated in a boycott, recall the experien~ and use the name of the boycott target below and treal it as if you
were boycotting that larget now.
Boycott Target ;
Regarding the boycott situation, please indicate how strongly you agree or disagree wllh the following statements.
strongly
disagree agree
I consider this issue a l-~ /"\
-f) Csocial injustice, .~--
This situation is unfaIr {, rt 0 }-.~forthose affected, " ,,~-
i baileve the
company's actions are {) -,"".- (1 ,-,
->"','- '->..j
egregious.
I regard the company's t-~ f; ~,,",,"
actions as unethical ' . L; o
/-'lo
.....~
..
.~ {""\
',,' ."",'
r, "'-':,
.'
"",,,->- (-':
'" .. j
115
How would you describe your alHtude towards [company]?
Unfavorable
Negative
,..,
",-" r,
Bad
/1 ("\
Favorable
r\ ;,-..~
"_,J
Positive
-',>-
Good
() ()
If you are participating in any other boycoJ1s or collective actions, please list or describe these other actiVities in
the box below.
In general, how would you characterize boycotts based on Ihe descriptions below
very bad Idea very good idea
fl··" r~ r-, () r~"";,....., ;"..-j
-'.'"
nol at aU useful very useful
\ ..-,,' ?~ "...., (,"",-"
nol at all favorable very favorable
0 C) 1""", r-~ 0
very negative very positive
r", ,.-, t-'~ ?-\ ,..-" (l / ..".,.
",,.' ;.. ..-.'
How much did each of the reason below motivale you 10 participate jn the boycott?
Thrill of victory
Punish the offending party
Altruism
Express my anger and displeasure
A need to take control of the situation
Clean hands - J did nol want anything to
do with the situation
Did not
moHvate
me at aU
C)
o
()
()
Motivated
mea
great
deal
116
If you had to identify one of the above reason as the primary motivation to boycott, II would be"
r. A need to tilke control 01 the situation
Altruism
Thnll Q1 victory
Punish the offending party
Express my anger and displeasure
Clean hands ~ I did not want: anything to do with the situation
Emotions
Which of the following emotions (either towards [company) or as part of the boycott situation) motivates
you to boycott?
Not at all
Trust
Suspicion
Anger
Betrayal
Not at all
Hatred
Hope
Guilt
Disgust
Shame
Contempt
Embarrassment
Regret
Cost-Benefits
,-.
Not at all
i"\
n
v
,<'"',! !-.~ .....
'.,,-;
"'"
"., ..-'~
\j
.'"' ~""-
'".'"
"', .r,
"
Please indicate below your agreement or disagreement with each of the statements regarding your boycott
against [company].
Everyone should take
part in the boycoll
because every
contribution, no matter
Strongly
Disagree
G
Strongly.
Agree
how small. is
important
Boycotts are an
effective means to
make a company
change its actions.
By boycotting, I can
help change the
actions of [company]
Strongly
Disagree
!'''\
117
Strongly
Agree
Please indicate below your agreement or disagreement with each of the statements regarding your boycott
against (company].
I would feel guilty if I
bought lcompanyj's
products.
I would feel
uncomfortable If other
people who are
boycotling saw me
purchase or consume
/companyJ's products.
My friends/family are
encouraging me 10
boycotl [company]
I will feel better about
myself If I boycott
[company].
.""'.
Strongly
Agree
C)
Please state how strongly you agree or disagree with the following statements.
Strongly
Disagree
Strongly
Agree
I probably do not buy
enough [company]
products for it 10 be
worthwhile
boycotting: It would
not even be noticed.
! probably do not
need to boycott
[company]: enough
other people are
doing so.
()
o
o
Boycolling may be
counterproductive to
negotiating a
agreeable
compromise with
[company).
Boycotting may put
Icompanyj's workers
in danger or cause
unforeseen harm to
those who are not
responsible for the
situation.
Strongly
Disagree
() ()
118
Strongly
Agree
Please state how strongly you agree or disagree with the following statements.
Strongly
Disagree
Strongly
Agree
~-. (.~ (:". /""'1, (: r'1 r, ()
It would be difficult for
me to give up the
produet(s) from
[company]
It would be·difficult for
me to give up the
produet(s)
Qualitative
t) f-"~ r'l. r, ,.-.,'~",-- \'-/ ,;
We would like you to tell us more aboul your boyeoll against [companyl. For each of Ihe following questions,
please feel free to discuss feelings, thoughts, reasons, or any other factors you feel are aruclalto your boycott
participation, .
What is the story behind what motivated you to join the boycott against [company)? How did you find out about
ttle boycott?
Under what circumstances would you end your boycoil of [company]?
00 youleel morally superior to your peers that do nol boycott? Explain why or why no!.
How would you describe the current success/efforts of your fellow boycott participants?
Individual Difference
Finally, we would like to know a bil about you
Please rate how strongly you agree or disagree with each of the following statements.
Strongly ,Somewhat Somewhat
Disagree Disagree Disagree Agree Agree
119
Strongly
Agree
One should be kind to all people
One should find ways to help others Jess fortunate
than oneself
A person should be concerned about the well-being of
others,
There should be equality for everyone--because we
are all human beings.
Those who are unable to provIde for their basic needs
should be helped by others.
A good society IS one in which people feel responsible
for one another
Everyone should have an equal chance and an equal
say in most things
Acting 10 prolect lhe rights and interests of other
members of the community is a major obligation for all
persons.
0 ()
0 n'",'
() (j
fJ (1
( ) ()
t: % ()
() 0
0
o
o
C)
t"'\
()
()
Strongly Somewhat Somewhat
Disagree Disagree Disagree Agree Agree
120
Strongly
Agree
In dealing with criminals the courts should recognize
that many are victims of circumstances,
Prosperous nations have a moral obligation to share
some of their wealth with poor nations,
Attitudes
Are you a member of a union?
c~ Yes
C No
()
r, fJ
Unions are ac.tive boycott participants and we are interested in your opinions relating to unions. Please indicate
below your agreement or disagreement with each of the statements
1 Strongly Somewhat Somewhat Strongly
Dis8wee Disagree Disagree Agree Agree Agree
Labor unions should be regulated
to a greater extent by the (j f"' rJ to: ,......\'-.. ,' <~-
government
We need more laws to limit the ,-. {?"', l"\ r\ (-1 ~-'1 i~ .~-~
power of unions. ~j \.-;' \.j '<../
The growlh of unions has made
r) f-~_ ~.- -,""'~ ~-~ () r~ () <'\
our democracy stronger \. .
Employees should have
representatives on Ihe board of
'-", ..
,-., n 1'.-....
directors of the organization for "-) ".~; ",,': ,~~.'
which the,' work.
Employe~s of an organization halle
better wages and working n '''' n (.~\ t~ [> C) ,. t""<conditions when all of Ihem belong ,...; ~,j ~, ...i
to a union,
The high wage demands of unions t"~ f-'~ r~~ r, r< C 0 Ct i~contribute directly to inflation. ;'j ~,.;.
During a strike, management
should be prohIbited by law from ('~ ~.~ "'--1, t') n n
hiring workers to take the place of "--/ '" ' ,,~
strikers
Workers should have Ihe light to
refuse to work in conditions which ~,...." '-, () C) t-",'~.J ,.j '~.,
they consider to be unsafe.
If the majority of workers in a
workplace vole to have a union, ~-~ ,", ,-, t"'>:. r", () i~ C~ f"">",the olhers should be required to V,F ',j <"",<
join.
Strongly Somewhat Somewhat Strongly
Disagree Disagree Disagree Agree Agree Agree
n c: ;,j rl (-:~ '":;,....;
----------------~---------
Unions impose too many
restr\c1ions on employers.
The selfishness of employers can
be fought only by strong unions
Workers should have to jOin a
union in order to hold a job.
...,
\./
c ()
121
Please mdicate below your agreement or disagreement with the following statement
Strongly Somewhat
Disagree Disagree Disagree Neutral
Big business has too much power in this
country
Demographies
What is your age?
What is the highest level of education you have completed?
•.., Less than High School
C) High Schooll GED
Soma College
C 4-yeal College Oeg",,,
Master's Degree
l-~ Doctoral Degree
Pm/essional Degree (JD. MD)
Wha1 is your gender?
Somewhat
Agree
c;
Agree
Strongly
Agree
What is your race?
White/Caucasian
f" Black/Alrican American
(-" Hispanic
male female
122
() Native American
Pacific lslander
Middle Eastern
Other
Please type below your email address if you are interesting in following this research and receiving a report of the
results? Our learn is very inlerested in keeping this research project transparent and accessible by the public.
Thank you for your participation I
APPENDIXC
STUDY 2 SURVEY INSTRUMENT & STIMULI
123
124
Exhibit A. Study 2 Survey Instrument (Stages 1 & 3)
The survey instrument for the first and third stages is identical. Each stage evaluates
participant attitude toward a variety of national brands then asks participants to choose
between these brands in a shopping simulation. At the end of the first stage, the
individual difference measures for humanitarian-egalitarian orientation and negative
attitude towards big businesses are also gathered.
Brand Evaluation Instructions (Stages 1 & 3)
PART 1. NATIONAL BRANDS
INSTRUCTIONS: You will be shown 16 national brands and be asked to rate your attitudes
towards these brands. You will also be asked about often you purchase products from these
national brands. Please move qUickly through each of the brands. There are no right or wrong
answers. We are primarily interested in people's opinions on common brands that you see
everyday at the shopping market.
After the instructions, participants are presented brand logos and corresponding brand
attitude measures. The following table summarizes the product category and brands
presented in the first and third stages of the study. The product categories marked with
an asterisk (*) are boycott targets in the second stage of the study.
Stage 1
Product Brands
Batteries* Duracell, Energizer
Chocolates* Hershey's, Nestle
Toothpaste* Colgate, Crest
Cola Coca-Cola, Pepsi
Detergent Purex, Tide
Ibuprofen Advil, Motrin
Frozen Pizza Freschetta, Totino's
St 3age
Product Brands
Batteries* Duracell, Energizer
Chocolates* Hershey's, Nestle
Toothpaste* Colgate, Crest
Cereal General Mills, Kelloggs
Shampoo Aveeno, Neutrogena
Cookies Famous Amos, Keebler
Sponges OCellO, Scotch-Brite
125
First, participants are asked to provide attitude measures for all 16 brands that correspond
to the above tables. An example of one brand attitude measure is presented as follows.
BRAND X
How would you describe your attitude towards ABOVE bra,nd?
Unfavorable
(;
Negative
Bad
f-""><;
'j
If the product from the ABOVE brand was not available when you are shopping,
never
>" how likely "','Quid you consider buying the
same productlrom another brand?
Favorable
Positive
Good
very
likely
Second, after all brands have been evaluated, the participants are asked to complete an
online shopping simulation based on the brands just evaluated.
Sho
PART 2. SHOPPING SIMULATION
Instructions: The next part of the study is a shopping simulation that investigates your familiarity
and preference for major U.S. brands. You will be asked to select between major brands and
identify how much you would be willing to pay for items on a shopping list. Do not worry if you
are not familiar with the brand, have never purchased the brand before, or have never used the
product before: simply choose a brand that appeals to you from the options given. Also, do not
worry about the providing exact estimates of the actual of the product. Just indicate how much
I you are willing to pay for the branded products based on personal shopping experience.
I
-----------
126
Participants are presented seven product categories in random order and asked to choose
one of the brands to put into their shopping cart. Below is an example of how each
product category is presented in the shopping simulation.
~ bottle of cola (or diet cQla) is on your shopping Ust. The following two brands are the only brands
available in the store. Which brand would you prefer to purchase?
Coca Cola PepsI Do not purchase
Use the slider bar below to indicate the maximum amount of money you would be willing to pay for a bottle of cola
of the brand you indicated.
$
o 2 4 6 8 12 14 16 18 20
--------------------------------
127
Exhibit B. Study 2 Survey Instrument (Stage 2)
Boycott Measures
Regarding the boycott situation, please indicate how strongly you agree or disagree with the following statements.
strongly
disagree agree
This situation is unfair 0 C)for those affected. -',,,"
I believe the
company's actions are 0 (-~ r. r-. /'''"~ ~...
egregious.
I consider this issue a
;', /-", -; r1
social injustice. ",£ \ "-../ "j
I regard the company's C) ,,""'> {", (;'
actions as unethical. \,-?'
Please indicate how likely you would engage in the following activities as a result of reading this editorial about the
company that Is the target of the boycott?
Discusses my dissatisfaction wjlh regards to the
company in privale
Consider and think about withholding purchase of
the company's products
Aclually withhold purchasing Ihe company's
products
Seek out and purchase a competitors products
Publicly discuss my dissallsfaction with Ihe
company
Publicly discuss boycoll and allemptto persuade
other people to join
Not
likely
r~
r~ C)
C) rJ
(j 0
o
Very
likely
(J
()
o
()
Based on what you have read, please indicate how likely you are to participate in the boycott mentioned in the
article:
Definitely not boycott Definitely boycott
.r"'~ ",...... f,\ i"\ () 1'-',,-~J
'.
,>".,i
Extremely unlikely Extremely likely
1'\ ~......~
""' C) "'J t}
N01 at all probable Highly probably
l""-.; n 0
While reading the news article based on the instructions gIven before the boycott article, please indicate the degree
128
to which you felt the following emotions?
Very
Not at all strongly
Trust r.....'< f"'~· '"'
("'s; ,-.,
'-..,.,:
':...../ :'-~/ ;-.,.d' '.j
Suspicion l-e\ ?~~ !~ t'"'\ r~~'.,..; 0;, .....
Anger n (", {-~
'""
Betrayal ?'""" r·,; r-' ...~>", ro,.
Hatred p ••.., ~..~': /'''- i-'"'\.\ ....
Hope r\ f-'";"'"-,
-
-'
Very
Not at all strongly
Worry r" l"\ i~ l""\ t~<A' ...,.;
Guill r'< ; ; r·,\
".. ~j
Happy ("~ t-1 ..... .-i ,-, r: "-., :;
Disgust ("0, r, r\ r, , r,,-:'~ ..J ""'~ \.i ",; ' ......'
Sadness I-~, <''''''': ,~-", ~-~
"j '-.j ,,<.F -~- ;
Shame r~ ~~~ r< ("',< "'
r,
~~..-/ <---' -'-'';') , ,:
Very
Not al all strongly
Contempt r""'...., r, C) n :,""",,! r~ f1l"...,; j
Fear t-,,- ~-~ (-'., (\ {-.. C}
Embarrassment ",Ole 1"-. r" f-., r.,',~;
Surprise r1
'"'
r--. I""". !"',
"'-.4 '<,/ ',,"- .~ ;
Happiness A n r, r\ c:",~,; '""...! >' ~j ,-<"",'
Regret r, :) r, !"', ,~ ()\-,-" ' ..' ';...1
Very
Not at all strongly
Compassion f" C) (\ ('\ ,-" ,-', ()',.H \)
Empathy f'''\ r, !-', !-~ l-", rl ;",\., '"j \,..i <~. ',.," " ....
Sympathy 7"--'" :r'""~ {-''', ('1 r~
'--W.-
Optimism C.J ,-." , , r~ ()
Remorse {""\
""'
{~'''' {"\ r,
'-'...J .,; .'<.;-:" ~x.~<
Very
Not at all strongly
Based on what you have just read, please indicate to What extent...
not at all
,..Do you consider
yourself responsible for
this situation?
()
extremely
.. ,Is the silllation
desirable to you?
,. 00 you consider the
company responsible
for this situation?
.Do you feel actions
you take can help the
situation?
..00 you care about
the Issue presented?
not al all
o
()
()
129
eldremely
Please Indicate below your agreement or disagreement with each of the statements regarding the boycott against
the company in the article.
Strongly
Agree
By boycotting, ! can help change the 0 f-~ r~ A () r~ () l"\
actions of the company
Boycotts are an effective means to
make the company change its t-~ (1 0 r· et .~-\ ,-,
actions.
Everyone should take part in the
boycott because every contribution. I~'-' () 0 I' r"j .,,';
no mailer how small, is important
Please indicate below your agreement or disagreement With each of the statements regarding the boycott against
the company in the article.
Strongly Strongly
Disagree Agree
I will feel better about myself if I boycott
.-. r", C) CJ ,-, (;this company. ..
My friends/family would encorage me to
,-" 0 ~~ C) (--:"boycoillhis company, .,.~ '....,/ ,.h \.j
I would feel uncomfortable if other
people who are boycotting saw me Cl ~~', C) ,-, () .~-.,purchase or consume thiS company's ;"'c/
products
I would feel guilty if I bought this
,..-" ;-\ (~ ......\ "''''", C) (; n CJ
company's products. \j \.<) \-i: ..
Ptease state how strongly you agree or disagree with the following statements regarding the boycott against the
company In the article•.
Strongly
Disagree
Sirongly
Agree
130
Strongly Strongly
Disagree Agree
I probably do not need to boycott this
company; enough other people are () () f-~ .,co",\~J
doing so,
I probably do not bUy enough of this
company's products for it to be
0 c: (\ h I""~ ~-,
worthwhile boycolting; it would not ',j \)
even be noticed,
Boycolting may be counterproduclive
to negotiating a agreeable compromise r; <-~ 0 t-., () () "-,, "'-. '-,';"--..'; :",.!
with this company,
Boycotting may put this company's
workers in danger or cause unforeseen
,,"""; ,--~ (1 r-' () r-~ (I r-~ (-'!: ,,-~harm to those who are not responsible ">~<'; ,~.- ';'../ ',.~,'
for the situation.
Please state how strongly you agree or disagree with the following statements regarding your purchase and
consumption of the boycott product.
Strongly
Disagree
Strongly
Agree
II would be difficult for me to give up
the produd(s) from [company)
It would be difficull for me to give up
the product(s)
0 (1 T \ r>-. r\-~,.
C) ('l r~ C) ,-\ i-"""j
".".-,' '-...,-/
Strongly
AgreeAgree
Somewhat
Agree
/"'~ r'} .~,
0 ~-. t"": ()..._,i
() r.. f-'" ()
-... .J ."j'
:~()
f)
Strongly Somewhat
Disagree Disagree Disagree
ThiS boycott is a very effective way of getting
the company to stop its business practices
I am confident that the consumer boycott will
ensure that the company will cease its
actions
I am certaIn that boycoll action will force the
company to make reparations to the affected
parties
Please indicate below if you agree or disagree regarding the effectiveness of consumer boycotts.
Neither
Agree
nor
Disagree
How much or how little do you care about the issue described in the articte?
Do not care at afl Care a great deal
()
131
To what extent are you concerned about human rights in developing countries?
Not at all concerned
How bothered are you by the human rights violation reported?
Not bothered at all
~-...
Extremely concerned
Extremely bothered
How Important or unimportant Is It to you that company ceases Its business practices as reported In the article?
Not at all important
()
tn general, how would you characteriz.e boycotts based on the descriptions below:
not at all useful
Extremely Important
very useful
very bad idea
very negative
not at all favorable
Manipulation Check· BOYCOTT
o
()
n
()
()
o
very good idea
very positive
very favorable
In the article about the consumer boycott, what product did the article want you to boycott?
Chocolates
8atteries
CJ Toothpaste
,., Detergent
In the article about the consumer boycott, what brand did the article want you to boycott? (Please type the name of
the brand or company in the space provided below),
132
Exhibit C. Manipulation (Perspective)
Note: These were the instructions provided to participants as part of the perspective
manipulation. Each participant was shown one of the following instructions prior to
reading the boycott stimuli.
Emotional
Rational
Control
We are interested in your feelings and emotions regarding the topic in
the following news article. Please read the article from a very
involved perspective, paying special attention to your feelings and
how you feel about the situation. It is VERY IMPORTANT that you
focus on your emotions rather than think about the economic
implications or the pros and cons of the situation.
We are interested in your thoughts and opinions regarding the topic
in the following news article. Please read the article from a detached,
objective, and rational perspective and consider the economic impact
on you. It is VERY IMPORTANT that you focus on the economic
implications and pros and cons ofthe situation and that you ignore
your feelings regarding the situation.
We are interested in your feelings and thoughts regarding the topic in
the following news article. Please read the article carefully and
answer the following questions.
133
Exhibit D. Toothpaste Boycott - Positive Frame (i.e. Crest)
Free Burma Coalition Calls for Proctor & Gamble Boycott
By Steven J. Brown
WASHINGTON, D.C. - Over the last
decade, most prominent American
companies have pulled out of Burma in
response to the mounting pressure
against that country's widespread and
grave human-rights violations. One
company that has so far resisted such
pressures is the consumer goods giant
Proctor & Gamble, which continues to
sell its products in Burma. However,
that may change with the recent call for
a consumer boycott of all P&G products
including popular toothpaste brand
Crest.
In May 1990, the people of Burma
democratically elected the party of
Nobel Peace Laureate San Suu Kyi to
government by an overwhelming 82% of
the vote. Soon after this landslide
victory however, Prime Minister Suu
Kyi was imprisoned and her party
crushed by a group of powerful generals
called SLORC (State Law and Order
Restoration Council). To this day,
SLORC maintains its reign of terror,
enforcing its will by rape, forced child
labor, and torture. This group continues
to torture Suu Kyi's supporters, attack
and displace millions of civilians, and
enslave thousands of children as army
porters and human mine sweepers.
(Original text modified from Sen et al. 2001)
The Free Burma Coalition issued a
statement yesterday asking all
consumers to boycott P&G products,
including Crest branded toothpaste, until
the company pulls out of Burma.
Calling on all consumers to use their
purchasing power to put a stop to the
pervasive violation of human rights in
Burma, Kathy Jo Benton, spokesperson
for this human rights organization,
emphasized the importance of consumer
boycotts in weakening the military's grip
on this small south-east Asian country.
Speaking on behalf of The Free Burma
Coalition, Ms. Benton stated that "All
foreign investments are regulated by
SLORC. It is not possible to do business
in Burma without directly supporting the
military government and its
unacceptable violation of human rights.
The simple act of refusing to buy P&G
products, like Crest branded toothpaste,
will send a clear message to the
company that responsible consumers
will not abide by corporate sponsorship
of an authoritarian regime, demonstrate
to the Burmese people that the world is
aware of their plight and willing to take
action, and help end the brutality and
slavery in Burma."
----------------- -- ------ ...-----
134
Exhibit E. Toothpaste Boycott - Negative Frame (i.e. Crest)
Free Burma Coalition Calls for Proctor & Gamble Boycott
By Steven J. Brown
WASHINGTON, D.C. - Over the last
decade, most prominent American
companies have pulled out of Burma in
response to the mounting pressure
against that country's widespread and
grave human-rights violations. One
company that has so far resisted such
pressures is the consumer goods giant
Proctor & Gamble, which continues to
sell its products in Burma. However,
that may change with the recent call for
a consumer boycott of all P&G products
including popular toothpaste brand
Crest.
In May 1990, the people of Burma
democratically elected the party of
Nobel Peace Laureate San Suu Kyi to
government by an overwhelming 82% of
the vote. Soon after this landslide
victory however, Prime Minister Suu
Kyi was imprisoned and her party
crushed by a group of powerful generals
called SLORC (State Law and Order
Restoration Council). To this day,
SLORC maintains its reign of terror,
enforcing its will by rape, forced child
labor, and torture. This group continues
to torture Suu Kyi's supporters, attack
and displace millions of civilians, and
(Original text modified from Sen et aI., 2001)
enslave thousands of children as army
porters and human mine sweepers.
The Free Burma Coalition issued a
statement yesterday asking all
consumers to boycott P&G products,
including Crest branded toothpaste, until
the company pulls out of Burma.
Calling on all consumers to use their
purchasing power to put a stop to the
pervasive violation of human rights in
Burma, Kathy Jo Benton, spokesperson
for this human rights organization,
emphasized the importance of consumer
boycotts in weakening the military's grip
on this small south-east Asian country.
Speaking on behalf of The Free Burma
Coalition, Ms. Benton stated that "All
foreign investments are regulated by
SLORC. It is not possible to do business
in Burma without directly supporting the
military government and its
unacceptable violation of human rights.
If consumers continue to buy P&G
products, like Crest branded toothpaste,
the company will never be held
accountable for its socially irresponsible
actions, consumers will fail to
demonstrate their support for Burmese
people, and the brutality and slavery in
Burma will never end."
135
Exhibit F. Chocolate Boycott - Positive Frame (i.e. Hershey's)
Global Exchange Calls for Hershey's Boycott
By Steven J. Brown
WASHINGTON, D.C. - Over the last
decade, most prominent American
confectionaries have stopped using slave
cocoa beans and moved to using Fair
Trade cocoa beans in their products in
response to mounting pressures against
the widespread human-rights violation in
the harvest of cocoa beans. One
company that has so far resisted such
pressures is Hershey's which continues
to purchase cacao bean from sources that
employ child slaves. However, that may
change with the recent call for a
consumer boycott of all Hershey's
products.
In an in-depth BBC report, it is
estimated that at least 12,000 children
have been sold into slavery to provide
the necessary cheap labor needed to
harvest cocoa beans in the Ivory Coast.
The Ivory Coast produces over 80% of
the beans sold by the global cocoa
exchange. Boys as young as 11 years
old from neighboring impoverished
countries of Benin, Togo, and Mali are
kidnapped and sold into slavery to work
on cocoa plantations. These child
laborers are imprisoned on the farms and
work in harsh conditions. Children as
small as age 8 are expected to harvest
and carry 20-pound cocoa sacks sixteen
hours a day. UNICEF is extremely
alarmed at the rate of mutilation and
death on these farms.
The Global Exchange issued a recent
statement asking all consumers to
boycott Hershey's products until the
company stops purchasing cacao beans
harvested by child slaves. Calling on all
consumers to use their purchasing power
to put a stop to the proliferation of
slavery, Kathy Jo Benton, spokesperson
for the Global Exchange, emphasized the
importance of consumer boycotts in
influencing corporate social
responsibility. Speaking on behalf of
The Global Exchange, Ms. Benton stated
"it is inconceivable that a modern
company is willing to employ slavery to
produce something that we buy every
day when alternatives like Fair Trade
chocolates are readily available. The
simple act of refusing to buy Hershey's
chocolate products will send a clear
message to the company that responsible
consumers will not abide by the gross
abuse of human rights for profit,
demonstrate to the child slaves that the
world is aware of their plight and willing
to take action, and help end the
proliferation of slavery."
136
Exhibit G. Chocolate Boycott - Negative Frame (i.e. Hershey's)
Global Exchange Calls for Hershey's Boycott
By Steven .T. Brown
WASHD\fGTON, D.C. - Over the last
decade, most prominent American
confectionaries have stopped using slave
cocoa beans and moved to using Fair
Trade cocoa beans in their products in
response to mounting pressures against
the widespread human-rights violation in
the harvest of cocoa beans. One
company that has so far resisted such
pressures is Hershey'S which continues
to purchase cacao bean from sources that
employ child slaves. However, that may
change with the recent call for a
consumer boycott of all Hershey's
products.
In an in-depth BBC report, it is
estimated that at least 12,000 children
have been sold into slavery to provide
the necessary cheap labor needed to
harvest cocoa beans in the Ivory Coast.
The Ivory Coast produces over 80% of
the beans sold by the global cocoa
exchange. Boys as young as 11 years
old from neighboring impoverished
countries of Benin, Togo, and Mali are
kidnapped and sold into slavery to work
on cocoa plantations. These child
laborers are imprisoned on the farms and
work in harsh conditions. Children as
small as age 8 are expected to harvest
and carry 20-pound cocoa sacks sixteen
hours a day. UNICEF is extremely
alarmed at the rate of mutilation and
death on these farms.
The Global Exchange issued a recent
statement asking all consumers to
boycott Hershey'S products until the
company stops purchasing cocao beans
harvested by child slaves. Calling on all
consumers to use their purchasing power
to put a stop to the proliferation of
slavery, Kathy .To Benton, spokesperson
for the Global Exchange, emphasized the
importance of consumer boycotts in
influencing corporate social
responsibility. Speaking on behalf of
The Global Exchange, Ms. Benton stated
"it is inconceivable that a modem
company is willing to employ slavery to
produce something that we buy every
day when alternatives like Fair Trade
chocolates are readily available. If
consumers continue to buy Hershey's
products, the company will never be
held accountable for its socially
irresponsible actions, consumers will fail
to demonstrate their support for child
slave, and the proliferation of slavery
will never end."
137
Exhibit H. Battery Boycott - Negative Frame (i.e. Duracell)
CWA Calls for Duracell Boycott
By Steven J. Brown
WASHINGTON, D.C. - Over the last
decade, most prominent American firms
have closed their overseas workshops
that required physical handling of spent
dry cell batteries to reclaim the
recyclable carbon cores. Most firms
now favor automated recycling facilities
both for safety and efficiency reasons.
However, Duracell continues contract
workshops in the poorest Asian
countries, such as Bengladesh, which
often skirt global child labor laws. This
may change with the recent call for a
consumer boycott of all Duracell
products.
A recent CNN documentary reported
that children as young as three or four
break open discarded batteries with
hammers in order to remove the
recyclable carbon rods and tiny pieces of
reusable metal. They earn 30 to 50 taka
(15 to 20 U.S. cents) per day. The
environment in and around the
workshops is polluted by carbon dust
and other toxic material. The hours are
long, the work tedious, and everything-
walls, ceilings and even the children's
faces-is covered with black carbon
dust. The children constantly lick their
lips to keep them wet, literally eating
toxic dust particles. Often only the
whites of their eyes and red shiny lips
are visible. The particles they inhale can
lead to black lung disease and cancer.
Child Workers in Asia (CWA), an
association of about 70 non-
governmental organizations and groups
in 14 countries, issued a recent statement
asking all consumers to boycott Duracell
products. Calling on all consumers to
use their purchasing power to put a stop
to the proliferation of child labor, Kathy
Jo Benton, spokesperson for the CWA,
emphasized the importance of consumer
boycotts in influencing corporate social
responsibility. Speaking on behalf of the
CWA, Ms. Benton stated "it is
inconceivable that a modern company is
willing to employ young children in such
hazardous conditions when mechanized
alternatives exist. The simple act of
refusing to buy Duracell batteries will
send a clear message to the company
that responsible consumers will not
abide by the gross abuse of human rights
for profit, demonstrate to the child
laborers that the world is aware of their
plight and willing to take action, and
help end to child labor in the world."
138
Exhibit I. Battery Boycott - Negative Frame (i.e. Duracell)
CWA Calls for Duracell Boycott
By Steven J. Brown
WASHINGTON, D.C. - Over the last
decade, most prominent American firms
have closed their overseas workshops
that required physical handling of spent
dry cell batteries to reclaim the
recyclable carbon cores. Most firms
now favor automated recycling facilities
both for safety and efficiency reasons.
However, Duracell continues contract
workshops in the poorest Asian
countries, such as Bangladesh, which
often skirt global child labor laws. This
may change with the recent call for a
consumer boycott of all Duracell
products.
A recent CNN documentary reported
that children as young as three or four
break open discarded batteries with
hammers in order to remove the
recyclable carbon rods and tiny pieces of
reusable metal. They earn 30 to 50 taka
(15 to 20 U.S. cents) per day. The
environment in and around the
workshops is polluted by carbon dust
and other toxic material. The hours are
long, the work tedious, and everything-
walls, ceilings and even the children's
faces-is covered with black carbon
dust. The children constantly lick their
lips to keep them wet, literally eating
toxic dust particles. Often only the
whites of their eyes and red shiny lips
are visible. The particles they inhale can
lead to black lung disease and cancer.
Child Workers in Asia (CWA), an
association of about 70 non-
governmental organizations and groups
in 14 countries, issued a recent statement
asking all consumers to boycott Duracell
products. Calling on all consumers to
use their purchasing power to put a stop
to the proliferation of child labor, Kathy
Jo Benton, spokesperson for the CWA,
emphasized the importance of consumer
boycotts in influencing corporate social
responsibility. Speaking on behalf of the
CWA, Ms. Benton stated "it is
inconceivable that a modern company is
willing to employ young children in such
hazardous conditions when mechanized
alternatives exist. If consumers continue
to buy Duracell batteries, the company
will never be held accountable for its
socially irresponsible actions, consumers
will fail to demonstrate their support for
child laborers, and the proliferation of
child labor will never end.
APPENDIXD
STUDY 3 SURVEY INSTRUMENT, PRO-BOYCOTT MESSAGE,
AND COUNTER-MESSAGES
139
140
Exhibit A. Study 3 Survey Instrument
CONSENT TO PARTICIPATE IN RESEARCH FORM
This research lS being conducted by Johnny Chen. a doctoral student althe University of Oregon. The goal of
this survey is Lo gauge consumer reactions towards chocolates and major chocolate brands in addition to positive
and negative message about these brands. This study has three sections;
I. Chocolates and Major Brands. Participants will be asked general qnestions and attitude about chocolate
purchases and specific brands.
II. Press Releases. Participants will he asked to provide reactions to hypothetical press releases regarding the
brand identifled in Section L
III Personal and Demo&raphics Questions. Participants will be asked SOl1W personal questions.
We encourage participants to complete this survey in one sitting (approximately 20 minutes). Pmiicipntioll in
this study is volulltmy. YonI' answers will remain confidential and will bc considered only in the aggregate with
no reference to specific individuals. Personal infonnation gathered will be deleted, and no records will be
maiutained after analysis. This survey has no affiliation with an identified brand or the choc.olate industry,
If yOH have any questions regarding this research, contact Johnny Chell at.jclteu3(@uoregol1.edu or (54 J)
346-1453, Ifyou have any questions regarding your rights as a research pmlicipulIl, please contact the Hnman
Su~iects Compliance Office at the University of Oregon (541) 346-2510.
SECTION I
Chocolates and Ma,jor Brands
How would you describe your general attitude towards chocolate?
Unenjoyable
Bad
Unpleasurable
Negative
Enjoyable
';.,.-"
~-~ 0
'"
Good
!'~ 0 r.,',"- "j
Pleasurable
~-< r .. l_~ ~
"""j
Positive
~-~. J'} c>
Unhealthy
Unfavorable
Dislike
Braud Hershey.
()
o
Before we start, how would yOll dl~scribe your altitude lowards Hershey's?
141
Healthy
Favorable
Like
Unfavorable
{'''~ C; ,-, (;'~ -' "'.'
Negative
(,1 ("'"> ()
Bad
/""'.
\j
Favorable
0 0
Positive
() r,,:
(iood
0 0
Please indicate if yOIl have personally purchased the following Hershey's produCIS. Please check alllhill apply.
CJAlmond Joy/Mounds
DHershey's Chocolate Milk Chocolate Bar
C) HerslJcy's Kisses
:"')Hershey's Bliss (any Iype)
.)Hershey's Pnt of Gold
:.JHcrshey's Skor
CJKit Kat (any Iype)
DMilk Duds
CJMr. Goodbar
'::iPay Day
URc:esc's Peanut Bulter Products (any type)
CJRolo
U Whatchamucall it
DWhoppers
'::JYork Peppermint Pattie
In tenns of quantity, how mallY oIthe products indicated above did you purchase in the lasl month? (Please lise
the slider to indicate the <Iuantily)
o
#..OfCh\:olate..f
5 10 15 20 25 30 35 40 45 50
products purchase last
month
Section U POST
o :5 10 15 20 25 30 35 40 45 50
142
Please refer to the text above and indicate hmv effective HershW's IJress release addresses the following points.
Communicate changes in Hershey's
practices
Communicate actions taken to make the
situation bettcr
Communicate tlul! Hershey's bas heard
the CQI1S11IllCrS' concern
Provide an explanation for the
Hershey's practices
Indicated that Hershey's has chosen the
best course of action
Indicated that Hershey's will not change
it actions
Convey negative repercussions of a
boycott
Explained the potential negat ive
outcomes of a boycolt
Attempt to make consumers feel good
about onese! f
Discuss the positive aspects of its brand
Discuss the positive aspects of its
produ<:ts
Very
Ineffective
o
()
c
o
Moderately
Effective
o
c)
()
o
()
()
t)
o
Very
Effe(~tive
Based on what you have read regarding Hcshey's, pl.ease indicate how strongly you agree or disagree with each
of the following statements.
I believe the
company's actions
are cgrcgioliS.
I regard the
company's actions as
disagree
o
'-.,
o
o c
strongly
ugree
()
unethicaL
This situation is
lUlfajr for ihose
affected.
I cOll5ader this issue a
sClcial injustice.
disagree
o
143
strongly
agree
,-. () ()
C! r-~ ,-'l .-, --~',,-. .;'
Please indicate how likely you would engage in the following activities as 11 result of reading the presented press
release(s) about Hershey's7
() ,., (-\
'J
r--~
"... ..
r." ,)<.-,,">
(J
--.
r. t}\ ..j- ;"",;
() 0 .~'.o!
Discuss my dissatisfaction with
regards to the company in
private
Consider and think about
withholding purchase oHhe
company's products
Actually withhold pnrclmsing
the company's products
Seek Ollt and purchase a
competitors products
Publicly discuss my
dissatisfaction \vith the
company
Publicly discuss boycott and
allclllpt to perslJarle other people
to joill
Vel)'
Unlikdy
()
o
CJ
t)
Very
Likely
C)
WillIe reading the press release(s), please indicate the degree to which yOIl felt the following emotions'?
Did not feel' Felt very
at all strongly
Hatred () :-~ - a ~~ r>;! -:;,.-,,"'
Sadness i""\ () f-'" () ()
Optimism " ~-~ C1 f: 0 ,-. ,-,t.; ).v~ "."" --:.,/
Anger (J -/'-, rJ ~--~ () n ,~-'~
-,.'- ';;-./
Compassion "-,,, j'"""", r .... (j\.j "._~ '''-d ~--,;-
Hope r~, ,-." l"\ t ....., 0 C (1,,~/ '-.,./ 0
Did not feel Felt velY
at all strougly
144
Did not feel Felt VCIY
at all slrongly
Shame C) (} f~ ,-~ \,f
"." v
Regret t""\ 0 'An""~--«';
Remorse t"~ F"", C,
'.- '-.,/
Em banassment t'-<,< 0 .r,~,-.j
Betrayal ?-'. "" r,'>..;
Disgust i""; r--;, () .r'''''",- t""\ r---.,
Did 110t feci Fdt velY
al all strongly
Wony r" () ,.,.,~ ,,"\, ;,.,~;
Sympathy (, 1'""'. C, () 0 '-',
Guilt r· (>: ~ () 0'., L'-
Contempt C'; r-"~ C) a c) .-, r~\,.; kJ
Empathy C1 ,..., ;;"'\ h:.,)
Did not feel Felt VCty
at all strongly
Based on what you have just read, please indicate to what extent..
".Is the situation
desirable to you?
, ,.Do yuu care abollt the
issue presented?
.. ,Do Y011 consider
yonrselfresponsible for
tll is situation?
...Do you feel actions
you take can help the
situation?
... Do you consider the
l:ompany responsible for
this situation?
not at all extremely
r·: ".""", ,..., r<' t""", ,,-... {-~
'<..d '~,.;/
,"->" ",-, ,.--, ,...,
'"' c:
"
':...,.) :'.,,-{
,-. !~ r~, r, 0
- 'd :,,.J
t-~ F" n "._\ /'. r.
l..... ..J <.J
-..
~
C)
Please indicate below your agreement or disagreement with each ofthe statements regarding the boycott against
the company in the article.
Boycotts are 1I11 effective means to
make the company change its
j Slmngiy
Disagree
() ()
Strongly
Agree
145
actions.
By boycotting, I can help change
the actions ofthe company.
Evelyone should take pali in the
boycott because every
contribution, no matter how small,
is important.
o
()
Stnlngly
Agree
Please indicate bela"" your agreemenl or disagreement \\lith each of the statements regarding IlIe boycott against
the company in the mticJe,
f-", f",,; ("\ t"'-~ [) ,r.., r, ! ...~ C~ ("-~'v ;"~,,
r--, r, ("\ ,-, [) ,.., l~ c: ~-) ,.....\\J ;,,'; \ ..,,' " ;, ;. ';.,/
I will leel betlcr about myself if I
boycotlthis eOlupany.
I would feel unCntIlI~lltable if other
people who are baycotling saw me
purchase or consume this company's
products,
My /i'iends/family would encourage
me to boycott this company.
I would feel guilty JJI hougllt Ihis
company's products.
Strong!y'
Disagree
() ()
Strongly:
Agree
Please slate how strongly you agree or disagree \vith the following statements regardill~ the boycott ap;ainst the
company in t11e article..
Strongly
:, Disagree
Strung!y'
Agree
I probably do not need to boycott
this company; enough other people
are doing so,
Boycotting may be
coulllcrprodm.:tive to negotiating an
agreeable compromise with this
company.
I probably do not buy enough of
this company's products for it to be
wortll'\vhile boycotliug; it would not.
even be noticed.
Boycotting may put this company's
workers in danger or cause
unforeseen harm to tlmsc who are
o
C: r, F) t-, c! C:) 0'''~..f. 'A","
,-~ ('" (1
'"
(J ~""'; 0 :r-,; G r~\j
r~ t-'lo. C\ () r. ,,-~
'"
() r·<) \j .,
"',,"
not responsible for the situation.
Strongly
Disagree.
.»•.•••...........
146
Strongly
Agree
Please state how strongly you agree or disagree with the following statements regarding your purchase and
consumption of chocolates.
It would be difficult for me to give
up chocolates
It would be d imcult for me to give
up chocolates from Hershey's ()
{"\
Strongly
Agree
Please draglhe end of tile slider bar to iudic<tte betow how su....~essfnl you feel the boycott will be. (0% =
failure; 100% total success)
o
. Boycott Ukelihoo..d of.1
Success
10 20 30 40 50 60 70 80 90 100
o
How would yon describe your attitude lowards your Hershey's after reading the press release(s)?
llnfuvorablc
Negative
Bad
CJ
Slimuli Manipulation Clleck - PR
Favorable
f~ (1 ,~
Positive
r, ..., (} r""""~,# >,.j
Good
"
C)
With regard to what you
.,. Think about allthc arguments
before making a decision 10 boycott
or not boycott?
... Go with your feeling.'i on whether
to boycott or not boycolt?
did you.,.
moderately
", 0 n C / .....,~cc.;
C1 () lN~, () c:
a great
deal
...Read the materials provided
carefully?
... Take the statements seriollsly?
no\ at
all moderately
147
a great
deal
Regarding Hershey's press release. how much do you agree with the [oHowing statements?
II Strongly Strongly
IIDisagree Agree
-nJe way this press release tries to persuade ('\ .~ r. {~~, :"\
people seems acceptable to me. \j
This press release tried 10 manipUlate the t';: ('\ t""",
audience in ways 1 do NOT Uke. ~. -~.j
1was annoyed by this press release beCalm:
it tried to inappropriately manage or control r; r>, C) (-~'~~~: ".J
my emotiOlls.
This press release tried to be persuasive
r. () f-' r. () ./..~ r~
without being excessively manipulative. -~.-.,'
The information presenred in this press
n r-\ n
rcleaSl' \vas fair and balant'ed. ,,/
To what extent do you agree that the statements made in Hershey's press release are:
Believable
Truthful
Re'illistic
Persuasive
Strong
Section III - Demographics
Strongly
Disag.ree
r,
:;.~
o
o
Section III
C1 /"\;,~/
r\ ........'"
:....,.; \j
C) r,
-,.j
0 0
r, 0
Strongly
Agree
You are "Imost done! This is the last !ie,tion nftlw survey and we are· Inter.:sted in getting to knnw ynu better. '.V" will be
asking fOl some basic demngruphics qllestions as wdl as asking for yom opinions.
Please rate how strongly you agree or disagl"cc with eacb of tbe following sll.1tements,
SU'ongly .Somewhat Smncwhat
Disagree Disagree Disagree Agree
Strongly'
Agree Agree
() ,,"-..., /'''\ t-~ t""; """'...
-,.-/
,-.. (~ r\ ,... C)L)
' .. ,,! \j
One should be kind to aU people.
One should find ways to help others less
fm1ullate Ilmn oneself
j\ person should be concerned about the
well-being of others.
There should be equality for evelyone--
because we are aU human beings.
Those who are unable to provide for their basic
needs should be helped by others.
A good society is one in which people feel
responsible for one another.
Everyone should have an eqnal chance and an
eqnal say in most things.
Acting to protect the rights lmd interests of
other members ofthe community is a major
obligation for all persons.
In dealing with criminals the courts should
recognize thaI many are victims of
circumstances.
Prosperous nations have a moral obligation to
share some of their wealth with poor nations.
Somewhat Somewhat
Disagree Agree Agree
Q c; r. r.',}
1'-", f'""';
.'"' C.'"" ""if \~j.
.'); ~~ .<-~
'... .J',..~; ",,-"; \,,..'
C) 0, ~-~ ,.-~.
"
,,; :>,,.;~ ,~,/;,.
Cl ,.-.-, /"\ r~
"
v =..".,' V
0 :~-.. f'"\. ~, C)....,-" '''.j i.: ...J
0 0 ,'-<:. t'""'.'.,-~) 0'
,,-..., r~ 1"\ t~
148
Strongly
Agree
o
~--;
Please indicate the level ofagreement or disagreement with the follmving statements.
. Strongly Somewhat
.Disagree. Disagree Disagree Neutral
Smlle\vhat
Agree Agree
Strongly
Agree
()
0 r~ !'-~ t] ,r,~,,-,,; -....j
If I want to be like someone, I often
try to buy the same brands that they
buy.
When buying products, I generally
purchase those brands that I think
others wiU approve of:
In Drder 10 insure that data is
collecting properly please select
"Agree" to the right.
I rarely purchase the latest fashion
styles until 1am sure my friel! ds
approve ofthem.
()
c; o
I achieve a sense ofbelongiag by
purchasing the same products and
brands that others purchase.
() 0 t-" ;' CJ ,\/ ",.
"
149
Somewhat Somewbat Strongly
Disagree Disagree Neutral Agree Agree Agree
Somewhat Somewhat Strongly
Disagree Disagree Neutral Agree Agree Agree
I fi'eqnenlly gather infonuation from
friends or l~IlUily about a product c; .,': {-~
be f()re 1 buy.
I often consult other people 10 help
choose the best alternative avaiIable () 0 l-' ())..,
trom <I product class.
1 like to know what brands and
products make good impressions on 0 t"••<~~;
others.
If other people can sec me using a
product, I often purchase the brand 0 () ~-~ l"'~ ?""'<: ~....\..,:
they expect me to buy.
Tt is important that others like the () r, F-'r) Cproducts and brands I buy. (j '"..; 'n'·-
Strongly Somewhat Somewhat Strongly
Disagree Disagree Disagree Neutral Agree Agree Agree
If I have little experience with a
product, I often ask my 1i'iends about (} ,,-<:: r", r, C"';
the product,
I often identify with other people by
purchasing the same products and r, ('1 r~, r, r\ r,:>...) '.,..J ,,,.£
brands they purchase.
To make sure I buy the right prodnct
or brand, I often observe what others L" i-, l-'" .,"'; (''''''
arc buying and using.
To what degree would you agree or disagree with the Jollowing statement:
; Strongly Somewhat
! Disagree Disagree Disagree
Big business has too IUuch power
in this country
What is your curren! age? (US. Census)
Less than 16
Co lolU 19
() 20!0 24
Neither
AflJ'ee
nor
Disagree
Somewhat
Agree Agree
o
Strongly
Agree
25 to 34
35 to 44
{~, 45 to 54
(~ 55 to 64
r, 65 years lind over
'Whal is your gender?
Male
,", femaLe
'What is your race?
White/Caucasian
'" Afhelln American
() Hispanic
Asian
Native American
() Pacific IslamicI'
,~ ,\rabie
Otber
What is the highest level of education yon have completed?
,-, Less than High School
C High School ,/ (JED
n Some College
C 2-year Col1ege Degree
~"; 4-year College Degree
n Master'g Degree
Doctoral Degree
(] Professional Degree (lD, MD)
150
151
Exhibit B. Pro-Boycott Message
BOYCOTT HERSHEY'S
The Center for Human Freedom (CHF) is encouraging consumers to boycott
Hershey's because it continues to sell chocolate products that include cacao, an
ingredient harvested with child slavery. Hershey's is the leading chocolate producer
in the U.S. controlling 42.5% of the market. Notable Hershey's product brands
include Hershey's, Reeses's, Milk Dud, Kit Kat, Scharffen Berger, and Dagoba. The
company's practices have changed little despite having signed legislation enacted in
2001 by Congress that was intended to forbid the use of cacao beans harvested by
child slaves. The only recourse left to American consumers is the complete
avoidance of Hershey's products.
To help you better understand the situation, we have summarized media reports from
familiar sources that have reported on the issue of slave chocolate. Also provided are the
links to the original stories.
CHILD SLAVERY
In 2000, the State Department's human rights report concluded that 12,000 children
between the ages of 9 and 12 were sold into forced labor on cocoa plantations on the
northem Ivory Coast. Forty-three percent of the world's cocoa beans, the raw material in
chocolate, originate from this country. A Knight-Ridder investigative report (Link) first
documented child slavery in the chocolate supply chain in 2001 and conditions have
changed little since then.
Imprisoned child laborers live and work in harsh conditions. The children subsist on com
paste and bananas. Thev are whinned. beaten and broken like horses in order to harvest
-""'- ." -.L -.L J
the almond-sized beans that are made into chocolate treats for more fortunate children
living in Europe and the United States. Children as young as age nine are expected to
harvest and carry 20-pound cocoa sacks sixteen hours a day. UNICEF is extremely
alarmed at the rate of mutilation and death on these farms.
152
(Exhibit B cont.)
A PERSISTENT PROBLEM
Hershey's signed the Harkin-Engel Protocol enacted by Congress in 2001 to eliminate
reliance on slave chocolates. However, CNN/Fortune magazine (Link,) reported that
Hershey's Chocolate continues to drag its feet in dealing with child slavery in its cocoa
supply chain. I'urthermore, the International Labor Rights Forum rated Hershey's at the
bottom in its 2009 Chocolate Scorecard (Jjnk) citing "the company has repeatedly
rejected shareholder resolutions related to protecting human rights and ensuring
transparency."
LAST OPTION
America is the world's largest chocolate consumer. According to Global Exchange
(Link), the US imports more than 729,000 tons ofcocoa beans/processed products, eats
more than 3.3 billion pounds of chocolate and spends more than $13 billion on chocolates
annually. People should be able to enjoy the simple pleasures of chocolate, but not at
such a high hwnan cost. As the industry leader, Hershey's should lead the change and
not stand idly by. Yet, as of this writing, Hershey's has effectively ignored public
sentiment as well as congressional legislation.
Therefore, the CHI' is initiating a nation wide boycott with its many partners starting in
2010.
WE HOPE THAT YOU WILL JOIN US.
153
Exhibit C. Original Counter-Message
HERSHEY, Pa. (BUSINESS WIRE) December 12,2009. - It has been some years since
media reports first appeared claiming widespread child labor abuses in West African
cacao farming. Much has happened since then.
At the time this issue was first raised, little was known beyond the fact that children
helped with the farm work, a typical practice in rural African communities. Now, we
have a better picture of the actual working conditions on these farms thanks to a
landmark, independent survey conducted by the International Institute for Tropical
Agriculture (IITA) in cooperation with the UN International Labor Organization (ILO)
and funded by The Hershey's Company.
Contrary to media reports, the survey found that the vast majority of farmers in the region
grow cacao responsibly, with no instances of forced child labor on the more than 4,500
farms included in the survey. The survey did identify several areas where change is
needed, including lack of access to education and farm safety issues involving machete
use and pesticide application.
A key element of this effort was the Harkin-Engel Protocol by the U.S. Congress that
established a series of date-specific steps intended to assure consumers that the chocolate
and cacao products they buy are not produced using the slave labor ofchildren. One of
the provisions of the 2001 bill includes the establishment of the International Cacao
Initiative Foundation whose sole mandate is to eliminate the worst forms of child forced
labor in cacao growing.
The final step under the Protocol timetable is the implementation of public certification
that cacao has been grown without the worst forms of child labor. This is the most
challenging step because it involves a certification process for a crop grown in a
developing country. Work on the design of this certification system, including
independent monitoring and verification as well as an international reporting process, is
already well underway.
These various efforts are aimed at improving the lives of millions of people who depend
on cacao farming for their livelihood, as well as for assuring consumers that the cacao in
their favorite chocolates have been grown responsibly. It will take time and the sustained
commitment of The Hershey Company and its industry partners to achieve meaningful
change. Much has already been achieved, but more improvement is expected as we move
forward to a time when all chocolate products are produced in accordance with
international labor standards and sustainable business practices.
154
Exhibit D. Acknowledge Action/Positive Change Counter-Message
HERSHEY, Pa. (BUSINESS WIRE) December 12,2009. -- It has been some years since
media reports first appeared claiming widespread child labor abuses in West African
cacao farming. Much has happened since then.
Hershey's fully acknowledges the untenable situation in West Africa. The company has
listened to customers and boycott advocates and has signed congressional legislation
aimed at curbing child labor abuses. Hershey's has also led the formation of the
International Cacao Initiative Foundation whose mandate is to eliminate the worst forms
of child labor and forced labor in cacao growing. In addition, Hershey's has contributed
portions of its profits and resources in an effort to increase monitoring of the region and
improve farming efficiency and incomes in the affected regions to reduce reliance on
child slavery.
A major problem has been in the supply chain where cacao bean wholesalers in West
Africa combine crops from numerous farms. This makes it impossible to separate out
cacao beans harvested with child labor from those harvested without such abuses. A
further complication is that intermediate processors typically blend various types of cacao
beans to produce chocolate that is designed for the American palate.
Hershey's is working as quickly as possible to reduce reliance on these tainted
chocolates. However, it will take the next decade to retool the entire supply chain so as
to produce sufficient certified products to serve the American market. Operations such as
Hershey's premium niche brand Dagoba have already adopted Fair Trade chocolate in its
products.
The Hershey Company recognizes that the chocolate and cacao industry must work
together on a pre-competitive basis to achieve progress in addressing the challenges and
opportunities facing the cacao sector. We ask that our loyal customers be patient while
Hershey's helps to change the structure of the chocolate industry.
We welcome consumer input and encourage consumers to be seek out products such as
Hershey's that are working towards improving sustainability and upholding good labor
practices.
155
Exhibit E. Reduce Efficacy Counter-Message
HERSHEY, Pa. (BUSINESS WIRE) December 12,2009. - It has been some years since
media reports first appeared claiming widespread child labor abuses in West African
cacao farming. Much has happened since then.
We have heard the concerns of consumers and boycott advocates in regard to the use of
forced labor in cacao growing. Sadly, this situation is unavoidable because it is part of
the global industry that Hershey'S works in. Cacao is a highly prized commodity that is
best sourced from certain regions of the world. The climate and cacao varieties in West
Africa produces the beans most sought after by industrial confectioners and American
consumers.
The global supply chain for cacao beans is an established institution that incorporates 1.5
million small farmers in West Africa, an extensive network of wholesalers and resellers,
and massive international networks of global food processors like ADM (Archer Daniels
Midland) and Cargill. Cacoa bean wholesalers in West Africa combine crops from
numerous farms. This makes it impossible to separate out cacao beans harvested with
child labor from those harvested without such abuses. In addition, there is no way to
monitor all the cacoa farms in the region. A further complication is that intermediate
processors typically blend various types of cacoa beans to produce chocolate that is
designed for the American palate. In order to produce the chocolate at competitive prices,
Hershey's has no choice but to continue acquiring its ingredients from these sources in
the interest of both American consumers and Hershey's shareholders.
Hershey's recognizes that the chocolate industry must work together on a cooperative
basis to achieve progress in addressing the challenges facing the cacao sector. Hershey's
will continue to work with the industry in an effort to improve working conditions. In the
meantime, Hershey'S will continue to produce chocolate for its primary market in the
United States.
156
Exhibit F. Negative Repercussions Counter-Message
HERSHEY, Pa. (BUSINESS WIRE) December 12,2009. - It has been some years since
media reports first appeared claiming widespread child labor abuses in West African
cacao farming. Much has happened since then.
The problem is much more complex than many boycott advocates and consumers
believe. Most people would conclude that it would be better to simply stop buying cacao
from West Africa, but experts say boycotting chocolate could make things worse for the
boys working on cacao farms. According to Anti-Slavery International, UNICEF, and
independent cacoa industry analysts, the price of cacao would drop dramatically if people
stopped buying chocolate. This means less money for everyone involved in cacao
production, especially the farmers. Farmers who use slave labor already say they do it
because they do not make enough money to pay the boys. If farmers earn even less
money, more boys will be forced to work for nothing.
It would be equally selfish for consumers to unilaterally decide on an action that has such
wide impact to the poor farmers working in West Africa - a region desperately poor with
only cacoa as a food crop that is in demand in the global economy. The climate and cacao
varieties in West Africa produce the most desirable beans for industrial confectioners and
American consumers. The American demand for chocolate has been good for these poor
West African nations. The stable prices for cacao beans improve the economy for 1.5
million small farms in this region.
So, while consumers may feel good about making a "clean hands" gesture, the sudden
and drastic reduction in chocolate consumption produced both by farmers using forced
labor and those who do not use such labor would have catastrophic consequences on local
economies. We therefore urge consumers to think more critically about the potential
negative consequences associated with any action they may take.
157
Exhibit G. Positive BrandlProducts Counter-Message
HERSHEY, Pa. (BUSINESS WIRE) December 12, 2009. - Hershey's is committed to
providing the highest quality chocolate for American consumers and is firmly committed
to making a difference in the communities where we live, work and do business both
locally and internationally. For more than 100 years, Hershey's has been the largest
producer of quality chocolate in North America and a global leader in chocolate and
sugar confectionery. In addition to its many iconic brands such as Hershey's Chocolate,
Hershey's Syrup, Hershey's Kisses, Kit Kat, and Reese's, Hershey's also produces the
super-premium brands Scharffen Berger and Dagoba. These ubiquitous brands bring
little pleasures to people and families' everyday.
Quality ingredients and world-renowned manufacturing practices have made Hershey's
the undisputed #1 chocolatier in the United States. Hershey's distinct chocolate flavor
starts with a single premium ingredient: West African cacao. The climate and cacao
varieties in West Africa produce the distinct chocolate flavor that has become an
American staple. Next, Hershey's manufacturing network transforms raw ingredients
into the world's best chocolate and confectionary products for consumers around the
globe. If one were to take a look at Hershey's sophisticated manufacturing facilities, it
would be indistinguishable from a modem silicon chip fabrication plant.
More than our products, Hershey's is proud of its legacy of giving back to the
community. Hershey employees volunteer actively, give generously and work to make a
positive difference where it is most needed. Both directly and through the United Way,
the company supports hundreds of community agencies that deliver services and support
to those most in need. The company's philanthropy reaches around the world, including
work with the Children's Miracle Network, Family Health International and a children's
bum center in Guadalajara, Mexico. Hershey's is a world leader in working to enhance
the lives of cacao farmers and their families in West Africa, Asia and the Americas. The
company supports programs that help improve farmer incomes, responsible labor
practices, opportunities for children and youth, and community health.
Hershey's practices environmental stewardship by supporting environmentally sound
cacao farming, implementing ongoing recycling, clean air and water management
programs, improving the environmental sustainability of our packaging and working to
reduce greenhouse gas emissions, waste, and our use of natural resources.
APPENDIXE
MEDIATION WITH CONTINUOUS MEDIATOR AND DICHOTOMOUS
OUTCOME (SPSS CODE)
158
159
/*THIS FILE WAS CREATED BY NATHANIEL R. HERR, FEBRUARY, 2006 */.
/*http://nrherr.bol.ucla.edu/Mediation/logmed.html */.
/*Gives you Standard Deviations for your variables*/.
DESCRIPTIVES
VARIABLES=xvar mvar yvar
/STATISTICS=MEAN STDDEV MIN MAX .
/*Gives you Covariance between X and M (top right or lower left box)*/
CORRELATIONS
/VARIABLES=xvar mvar
/PRINT=TWOTAIL NOSIG
/STATISTICS XPROD
/MISSING=PAIRWISE .
/*Gives you "a" coefficient and "a" coefficient standard error - Find
these in "Coefficients" box in output */
REGRESSION
/MISSING PAIRWISE
/STATISTICS COEFF OUTS R ANOVA
/CRITERIA=PIN(.05) POUT(.lO)
/NOORIGIN
/DEPENDENT mvar
/METHOD=ENTER xvar
/*Gives you "c" coefficient and "c" coefficient standard error - Find
these in "Variables in the Equation" box in output */
LOGISTIC REGRESSION VAR=yvar
/METHOD=ENTER xvar
/CRITERIA PIN(.05) POUT(.lO) ITERATE(20) CUT(.5) .
/*Gives you BOTH the "b" coefficient and the" c' " coefficient and
their standard errors - Find these in "Variables in the Equation" box
in output */
LOGISTIC REGRESSION VAR=yvar
/METHOD=ENTER xvar mvar
/CRITERIA PIN(.05) POUT(.lO) ITERATE(20) CUT(.5) .
160
Mediation with Dichotomous Variables
X=Causal variable a=path from X to M
M=Mediator variable b=path from M to Y (controlling for X)
Y=Outcome variabe c=direct path from X to Y
c'=path from X to Y (controlling for M)
L...-- I= you must input this information
Run descriptive statistics in SPSS for your variables for SDs
SD(X)= 1.11278 Var(X)= 1.238279
SD(M)= 1.53928 Var(IV1)= 2.369383
SD(Y)= 0.49851 Var(Y)= 0.248512
Run correlate with X and M variables and check "covariance matrix" box in options
COV(X,M)= I 0.585 1
Run regressions for continuous variables and logistic regressions for dichotomous outcome va
a= 0.81 SE(a)= 0.08
b= 0.264 SE(b)= 0.12
c= 0.382 SE(c)= 0.138
c'= 0.17 SE(c')= 0.166
Var(Y')= 3.470695 SD(Y')= 1.86298
Var(M')= 4.102435 SD(M')= 2.025447
Var(Y")= 3.543432 SD(Y")= 1.882401
comp a= 0.445014 SE(comp a)= 0.043952
comp b= 0.215879 SE(comp b)= 0.098127
comp c'=: 0.100495 SE(comp c')= 0.098131
comp c= 0.228173 SE(comp c)= 0.082429
ab+c'= 0.196564
= Enter these values into the Sobel Test
Aroian Sobel Goodman
Sab = 0.044894276 0.044687 0.044478
Sobel= 2.139892625 2.149836 2.1599191
FORMAT CREATED BY NATHAI\JlEL R. HERR, FEBRUARY, 2006
http://nrherr.bol.uda.edu/Mediation/logmed.html
APPENDIXF
STUDY 2 ML AND BAYESIAN PARAMETER ESTIMATES
161
162
Table F-1. Parameter Estimate (ML and Bayesian)
ML Bayesian
B SE I) C.R. B SE c.R.
NABB ----+ PMV ,095 .041 .133 2,30 .095 .042 2.26
HE ----+ PMV .263 .096 .166 2.75 .264 .101 2,61
PMV ----+ MEothcr ,776 .067 .583 11.50 .775 .069 11.23
PMV ----+ MEse'f .377 .086 .291 4.37 .379 .085 4.46
PMV ----+ BENE ,833 .119 .414 7.03 .833 .123 6.77
PMV ----+ COST -.160 .055 -.203 -2.93 -.160 .054 -2,96
MEOther ----+ BI .238 .067 .228 3.57 .239 .071 3.37
MEse1f ----+ BI .031 .054 .029 .57 .032 .057 .56
BENE ----+ BI .271 .042 .392 6.41 .270 .044 6.14
COST ----+ BI -.022 .085 -.012 -.25 -.021 .090 -.23
PMV ----+ BI .099 .099 .071 1.00 .098 .097 l.01
BI ----+ BOYCOTT .175 .020 .529 8.61 .175 .020 8.75
Note: HE = Humanitarian-Egalitarian Orientation; NABB = Negative Attitude towards
Big Businesses; MEOther = Other-condemning moral emotions; MESelt'= Self-conscious
moral emotions; BENE = Benefit evaluation; COST = Cost evaluation; BI = Boycott
Intentions; BOYCOTT = Boycott Behavior
163
Table F-2. Parameter Estimates (Bayesian) with Individual Discrete Emotions and Cost-
Benefit Factors
B SE f3 C.R.
Individual Differences
NABB ---+ PMV .091 .042 .131 2.17
HE ---+ PMV .265 .095 .162 2.79
Other-condemning Moral Emotions
PMV ---+ Anger .782 .094 .531 8.32
PMV ---+ Contempt .268 .104 .181 2.58
PMV ---+ Disgust .683 .096 .461 7.11
Anger ---+ BI .219 .063 .231 3.48
Contempt ---+ BI .048 .048 .047 1.00
Disgust ---+ BI .021 .056 .022 .38
Self-concious Moral Emotions
PMV ---+ Guilt .475 .119 .281 3.99
PMV ---+ Regret .410 .106 .271 3.87
PMV ---+ Shame .420 .107 .275 3.93
PMV ---+ EmbaLTassment .307 .104 .205 2.95
Guilt ---+ BI .010 .052 .005 .19
Regret ---+ BI .016 .060 .021 .27
Shame ---+ BI -.026 .072 -.030 -.36
Embarrassment ---+ BI .020 .055 .015 .36
Benefit
PMV ---+ Make a Difference .832 .153 .377 5.44
PMV ---+ Self-Enhancement 1.046 .132 .477 7.92
Make a DitIerence ---+ BI .052 .040 .081 1.30
Self-Enhancement ---+ BI .251 .053 .393 4.74
Cost
PMV ---+ Counterargument -.349 .104 -.231 -3.36
PMV ---+ Constrained Cons. -.030 .038 -.066 -.79
Counterargument ---+ BI -.004 .047 -.006 -.09
Constrained Cons. ---+ BI .. A A .140 r\ '. ,- -.31-.U,+,+ -.ut:)
PMV ---+ BI .054 .097 .038 .56
BI ~ BOYCOTT .135 .023 .417 5.87
APPENDIXG
STUDY 3 ML PARAMETER ESTIMATES
164
165
Table G-1. Parameter Estimates (ML)
Hypothesized B SE f3 C.R.
NA BB ----+ PMV .271 .030 .337 10.39
HE ----+ PMV .520 .050 .334 10.32
PMV
----+ MEOthcr .936 .042 .643 22.26
PMV
----+ MEscH" .822 .048 .552 17.16
PMV ----+ BENE .935 .047 .607 19.78
PMV ----+ COST -.136 .034 -.165 -3.97
MEOthcr ----+ B1 .086 .033 .080 2.64
MEscH' ----+ BT .350 .033 .332 10.77
BENE ----+ BI .498 .028 .488 17.82
COST ----+ B1 -.076 .028 -.040 -2.74
PMV ----+ BI .152 .034 .097 4.49
Non-Hypothesized B SE f3 c.R.
NABB ----+ BENE .091 .027 .073 3.33
HE
----+ MEOthcr .353 .065 .156 5.40
HE
----+ MEscH' .384 .075 .166 5.15
HE ----+ BENE .304 .071 .127 4.25
HE ----+ COST .135 .053 .105 2.53
M1 ----+ PMV -.632 .114 -.190 -5.56
M2 ----+ PMV -.507 .113 -.155 -4.51
M4 ----+ PMV -.462 .115 -.138 -4.03
M5 ----+ PMV -.351 .115 -.104 -3.06
Note: HE = Humanitarian-Egalitarian Orientation; NABB = Negative Attitude towards
Big Businesses; MEOther = Other-condemning moral emotions; MEselF Self-conscious
moral emotions; BENE = Benefit evaluation; COST = Cost evaluation; BI = Boycott
Intentions. Counter-Message Tactics: Ml = Original; M2 = Acknowledge Tactic; M4 =
Negative Repercussion Tactic; M5 = Positive Aspect Tactic.
166
Table G-2. Parameter Estimates (IVIL) with Individual Discrete Emotions and Cost-
Benefit Factors
B SE f3 p
Individual Differences
NA BB ~PMV .270 .050 .337 <.001
HE ~ PMV .520 .026 .334 < .001
Other-condemning Moral Emotions
PMV ~ Anger .985 .050 .598 < J)Ol
PMV ~ Contempt .946 .049 .607 <.001
PMV ~ Disgust 1.067 .043 .681 < .001
Anger ~ Boycott Intentions .072 .027 .076 <.001
Contempt ~ Boycott Intentions .053 .024 .052 <.001
Disgust ~ Boycott Intentions .017 .027 .076 I1S
Self-concious Moral Emotions
PMV ~ Guilt .976 .049 .601 <.001
PMV ~ Regret .888 .049 .563 <.001
PMV ~ Shame .914 .052 .553 <.001
PMV ~ Embarrasment .962 .044 .586 <.001
Guilt ~ Boycott Intentions .126 .028 .130 <.001
Regret ~ Boycott Intentions .092 .024 .092 <.001
Shame ~ Boycott Intentions -.020 .026 -.021 ns
Embarrasment ~ Boycott Intentions .065 .029 .068 <.05
Benefit
PMV ~ Make a Difference .987 .044 .643 < .001
PMV ~ Self-Enhancement 1.163 .045 .694 <.001
Make a Difference ~ Boycott Intentions .113 .027 .111 <.001
Self-Enhancement ~ Boycott Intentions .402 .031 .429 <.001
Cost
PMV ~ Counterargument -.176 .042 -.156 <.001
PMV ~ Constrained Cons. -.017 .042 -.015 ns
Counterargument ~ Boycott Intentions -.072 .019 -.052 <.001
Constrained Cons. ~ Boycott Intentions -.005 .019 .010 I1S
PMV ~ Boycott Intentions .131 .032 .084 <.001
167
BIBLIOGRAPHY
Aaker, J., Fournier, S., & Brasel, S. A. (2004). When Good Brands Do Bad. Journal of
Consumer Research, 31(1), 1-16.
Akaike, H. (1987). Factor analysis and AIC. Psychometrika, 52(3),317-332.
Anderson, D. R, Burnham, K. P., & White, G. C. (1998). Comparison of Akaike
information criterion and consistent Akaike information criterion for model
selection and statistical inference from capture-recapture studies. Journal of
Applied Statistics, 25(2),263 - 282.
Bagozzi, R P., Gopinath, M., & Nyer, P. U. (1999). The Role ofEmotions in Marketing.
. Journal ofthe Academy ofMarketing Science, 27(2), 184-206.
Bagozzi, R P., & Moore, D. 1. (1994). Public Service Advertisements: Emotions and
Empathy Guide Prosocial Behavior. Journal ofMarketing, 58(1), 56.
Baron, R M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in
social psychological research: Conceptual, strategic, and statistical considerations.
Journal ofPersonality and Social Psychology., 51(6), 1173-1182.
Batson, C. D. (1998). Altruism and Prosocial Behavior. In D. Gilbert, S. Fiske & G.
Lindzey (Eds.), The Handbook ofSocial Psychology (4th ed., Vol. 2). New York:
McGraw-Hill.
Beder, S. (2002). Nike's Greenwashing Sweatshop Labor. Retrieved July 4, 2010, from
http://www.organicconsumers.org/c1othes/nikesweatshop.cfm
Be1k, R. W. (1988). Possessions and the Extended Self. Journal ofConsumer Research,
15(2),88.
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological
Bulletin, 107(2),238-246.
Brown, T. J., & Dacin, P. A. (1997). The Company and the Product: Corporate
Associations and Consumer Product Responses. Journal ofMarketing, 61 (1), 68-
84.
168
Byrne, B. M. (2010). Structural Equation Modeling with AMOS: Basic Concepts,
Applications, and Programming (2nd ed.). New York: Routledge.
Cacioppo, J. T., & Petty, R. E. (1984). The Elaboration Likelihood Model of Persuasion.
Advances in Consumer Research, 11(1),673-675.
Chavis, L., & Leslie, P. (2009). Consumer boycotts: The impact of the Iraq war on
French wine sales in the U.S. [10.l007/s11129-008-9043-y]. Quantitative
Marketing and Economics, 7(1), 37-67.
Cialdini, R. B. (1998). Injluence: The Psychology ofPersuasion. New York: Quill.
Cooper-Martin, E., & Holbrook, M. B. (1993). Ethical Consumption Experiences and
Ethical Space. Advances in Consumer Research, 20( 1), 113-118.
Dickinson, R., & Hollander, S. C. (1991). Consumer votes. Journal ofBusiness
Research, 22(4),335-346.
Duhachek, A. (2005). Coping: A Multidimensional, Hierarchical Framework of
Responses to Stressful Consumption Episodes. The Journal ofConsumer
Research., 32(1),41.
Duhachek, A., & Iacobucci, D. (2005). Consumer Personality and Coping: Testing Rival
Theories of Process. Journal ofConsumer Psychology, 15(1), 52.
Duhachek, A., & Oakley, J. L. (2007). Mapping the Hierarchical Structure of Coping:
Unifying Empirical and Theoretical Perspectives. Journal ofConsumer
Psychology, 17(3),218-233.
Eisenberg, N., Miller, P. A., Schaller, M., Fabes, R. A., Fultz, J., Shell, R., et al. (1989).
The Role of Sympathy and Altruistic Personality Traits in Helping: A
Reexamination. Journal ofPersonality, 57(1),41-67.
Ekman, P. (1999). Basic emotions. In T. Dalgleish & M. Power (Eds.), Handbook of
cognition and emotion. Sussex, U.K.: John Wiley & Sons.
Friedman, M. (1985). Consumer Boycotts in the United States, 1970-1980:
Contemporary Events in Historical Perspective. The Journal ofConsumer Affairs,
19(1), 96-.
Friedman, M. (1991). Consumer Boycotts: A Conceptual Framework and Research
Agenda. Journal ofSocial Issues, 47(1), 149-168.
Friedman, M. (1996). Grassroots Groups Confront the Corporation: Contemporary
Strategies in Historical Perspective. Journal ofSocial Issues, 52(1), 153-167.
169
Friedman, M. (1999). Consumer Boycotts: Effecting Change through the Marketplace
and the Media. New York: Routledge.
Frijda, N. H. (1988). The laws of emotion. American Psychologist, 43(5), 349-358.
Gregoire, Y, & Fisher, R (2006). The effects of relationship quality on customer
retaliation. Marketing Letters, i7(1), 31-46.
Gregoire, Y, & Fisher, R 1. (2007). Customer betrayal and retaliation: when your best
customers become your worst enemies. Journal ofthe Academy ofMarketing
Science.
Haidt,1. (2003). The moral emotions. In R 1. Davidson, K. R. Scherer & H. H.
Goldsmith (Eds.), Handbook ofaffective sciences (pp. 852-870). Oxford: Oxford
University Press.
Healey, J. R (2010). Toyota delayed almost a year before issuing 2005 steering defect
recall. Retrieved May 29, 2010, from
http://www.usatoday.com/money/autos/2010-05-10-toyota-recall-de1ay_N.htm
Hirschman, A. O. (1970). Exit, Voice, and Loyalty: Responses to Decline in Firms,
Organizations, and States. Cambridge, MA: Harvard University Press.
Hirschman, E. c., & Holbrook, M. B. (1982). Hedonic Consumption: Emerging
Concepts, Methods and Propositions. Journal ofMarketing, 46(3),92-101.
Hoffman, M. L. (1987). The Contribution of Empathy to Justice and Moral Judgement. In
N. Eisenberg & J. Strayer (Eds.), Empathy and its Development (pp. 47-80). New
York: Cambridge University Press.
Hu, L.-T., & Bentler, P. M. (1995). Evaluating Model Fit. In R. H. Hoyle (Ed.),
Structural Equation Modeling: Concepts, issues, and Applications (pp. 76-99).
Thousand Oaks, CA: SAGE Publications, Inc.
Hu, L.-T., & Bentler, P. M. (1999). Cutoff Criteria for Fit Indexes in Covariance
Structure Analysis: Conventional Criteria Versus. Structural Equation Modeling,
6(1), 1.
Innes, R (2006). A Theory of Consumer Boycotts under Symmetric Information and
Imperfect Competition. Economic Journal, 116(511),355-381.
John, A., & Klein, J. (2003). The Boycott Puzzle: Consumer Motivations for Purchase
Sacrifice. Management Science, 49(9), 1196-1209.
Johnson, A. R, & Stewart, D. W. (2004). A Reappraisal of the Role of Emotion in
Consumer Behavior: Traditional and Contemporary Approaches. In N. K.
169
Friedman, M. (1999). Consumer Boycotts: Effecting Change through the Marketplace
and the Media. New York: Routledge.
Frijda, N. H. (1988). The laws of emotion. American Psychologist, 43(5),349-358.
Gregoire, Y, & Fisher, R. (2006). The effects of relationship quality on customer
retaliation. Marketing Letters, 17(1), 31-46.
Gregoire, Y, & Fisher, R. J. (2007). Customer betrayal and retaliation: when your best
customers become your worst enemies. Journal ofthe Academy ofMarketing
Science.
Haidt, J. (2003). The moral emotions. In R. J. Davidson, K. R. Scherer & H. H.
Goldsmith (Eds.), Handbook ofaffective sciences (pp. 852-870). Oxford: Oxford
University Press.
Healey, J. R. (2010). Toyota delayed almost a year before issuing 2005 steering defect
recall. Retrieved May 29, 2010, from
http://www.usatoday.com/money/autos/2010-05-10-toyota-recall-delay_N.htm
Hirschman, A. O. (1970). Exit, Voice, and Loyalty: Responses to Decline in Firms,
Organizations, and States. Cambridge, MA: Harvard University Press.
Hirschman, E. C., & Holbrook, M. B. (1982). Hedonic Consumption: Emerging
Concepts, Methods and Propositions. Journal o.fMarketing, 46(3),92-101.
Hoffman, M. L. (1987). The Contribution of Empathy to Justice and Moral Judgement. In
N. Eisenberg & J. Strayer (Eds.), Empathy and Its Development (pp. 47-80). New
York: Cambridge University Press.
Hu, L.-T., & Bentler, P. M. (1995). Evaluating Model Fit. In R. H. Hoyle (Ed.),
Structural Equation Modeling: Concepts, Issues, and Applications (pp. 76-99).
Thousand Oaks, CA: SAGE Publications, Inc.
Hu, L.-T., & Bentler, P. M. (1999). Cutoff Criteria for Fit Indexes in Covariance
Structure Analysis: Conventional Criteria Versus. Structural Equation Modeling,
6(1), 1.
Innes, R. (2006). A Theory of Consumer Boycotts under Symmetric Information and
Imperfect Competition. Economic Journal, 116(511), 355-381.
John, A., & Klein, J. (2003). The Boycott Puzzle: Consumer Motivations for Purchase
Sacrifice. Management Science, 49(9), 1196-1209.
170
Johnson, A. R., & Stewart, D. W. (2004). A Reappraisal of the Role of Emotion in
Consumer Behavior: Traditional and Contemporary Approaches. In N. K.
Malhotra (Ed.), Review ofMarketing Research (Vol. 1, pp. 3-33). Armonk, N.Y.:
M.E. Sharpe.
Kahneman, D. (2003). A perspective on judgment and choice: Mapping bounded
rationality. American Psychologist. (9), 697-720.
Katz, I., & Hass, R. G. (1988). Racial Ambivalence and American Value Conflict:
Correlational and Priming Studies of Dual Cognitive Structures. Journal of
Personality and Social Psychology, 55(6),893-905.
Klein,.1. G., Smith, N. C., & John, A. (2004). Why We Boycott: Consumer Motivations
for Boycott Participation. Journal ofMarketing, 68(3), 92-109.
Kozinets, R. V., & Handelman, J. (1998). Ensouling Consumption: A Netnographic
Exploration of The Meaning of Boycotting Behavior. Advances in Consumer
Research, 25(1),475-480.
Kozinets, R. V., & Handelman, J. M. (2004). Adversaries of Consumption: Consumer
Movements, Activism, and Ideology. Journal ofConsumer Research, 31(3),691-
704.
Lazarus, R. S. (1991). Progress on a Cognitive-Motivational-Relational Theory of
Emotion. American Psychologist, 46(8), 819-834.
Lazarus, R. S. (2001). Relational meaning and discrete emotions. In K. R. Scherer, A.
Schorr & T. Johnstone (Eds.), Appraisal Processes in Emotion (pp. 37-67). New
York: Oxford University Press.
MacInnis, D. 1., & de Mello, G. E. (2005). The Concept of Hope and Its Relevance to
Product Evaluation and Choice. Journal ofMarketing, 69(1), 1-14.
Mackinnon, D. P., & Dwyer, .1. H. (1993). Estimating Mediated Effects in Prevention
Studies. Eval Rev, 17(2), 144-158.
Madrigal, R. (2008). Hot vs. cold cognitions and consumers' reactions to sporting event
outcomes. Journal o/Consumer Psychology, 18(4),304-319.
Madrigal, R., LaBarge, M., & Chen, .1. (2007 ). A Hierarchical Model ofConsumers '
Response to Advocacy Advertising: Consumer Skepticism, Empathic Response,
Empathic Emotions, and Prosocial Behavior. Paper presented at the Society for
Consumer Psychology, Las Vegas, NV.
171
Mauro, R., Sato, K., & Tucker, 1. (1992). The Role of Appraisal in Human Emotions: A
Cross-Cultural Study. Journal ofPersonality & Social Psychology, 62(2),301-
317.
Miller, K. E., & Sturdivant, F. D. (1977). Consumer Responses to Socially Questionable
Corporate Behavior: An Empirical Test. Journal ofConsumer Research, 4(1), 1-
7.
O'Shaughnessy, 1., & O'Shaughnessy, N. 1. (2003). The Marketing Power ofEmotion.
New York, New York: Oxford University Press, Inc.
Philips, M. (2010). Getting Toyota Back On Track. from
http://www.newsweek.com/201 0102/09/getting-toyota-back-on-track.html
Pruitt, S. W., & Friedman, M. (1986). Determining the effectiveness of consumer
boycotts: A stock price analysis of their impact on corporate targets Journal of
Consumer Policy, 9(4),375-387.
Pruitt, S. W., Wei, K. C. 1., & White, R. E. (1988). The Impact of Union-Sponsored
Boycotts on the Stock Prices of Target Firms. Journal ofLabor Research, 9(3),
285-289.
Roseman, 1. J., & Smith, C. A. (2001). Appraisal Theory: Overview, Assumptions,
Varieties, Controversies. In K. R. Scherer, A. Schorr & T. Johnstone (Eds.),
Appraisal Processes in Emotion (pp. 3-19). New York: Oxford.
Rust, R. T., & Cooil, B. (1994). Reliability measures for qualitative data: Theory and
implications. Journal ofMarketing Research (JMR), 31(1), 1.
Scherer, K. R. (1997). Profiles of emotion-antecedent appraisal: Testing theoretical
predictions across cultures. Cognition & Emotion, 11(2), 113-150.
Scherer, K. R. (2001). Appraisal considered as a process of multilevel sequential
checking. In K. R. Scherer, A. Schorr & T. Johnstone (Eds.), Appraisal Processes
in Emotion (pp. 92-120). New York: Oxford University Press.
Sen, S., & Bhattacharya, C. B. (2001). Does Doing Good Always Lead to Doing Better?
Consumer Reactions to Corporate Social Responsibility. Journal ofMarketing
Research (JMR), 38(2),225-243.
Sen, S., Gurhan-Canli, Z", & Morwitz, V. (2001). Witb..holding Consumption: A Social
Dilemma Perspective on Consumer Boycotts. Journal ofConsumer Research,
28(3),399.
172
Shiv, B., & Fedorikhin, A. (1999). Heart and Mind in Conflict: The Interplay of Affect
and Cognition in Consumer Decision Making. [Article]. Journal ofConsumer
Research, 26(3),278-292.
Smith, C. A, & Ellsworth, P. C. (1985). Patterns of Cognitive Appraisal in Emotion.
Journal ofPersonality and Social Psychology, 48(4),813-838.
Smith, C. A., Haynes, K. N., Pope, L. K., & Lazarus, R. S. (1993). In Search of the "Hot"
Cognitions: Attributions, Appraisals, and Their Relation to Emotion. Journal of
Personality & Social Psychology, 65(5),916-929.
Smith, C. A, & Lazarus, R. S. (1993). Appraisal Components, Core Relational Themes,
and the Emotions. Cognition & Emotion, 7(3/4), 233-269.
Smith, N. C. (1990). Morality and the Market: Consumer Pressure for Corporate
Accountability. London: Routledge.
Smith, N. C. (2005). Consumer Activism: Boycotts, Brands and Marketing
Communications. Advances in Consumer Research, 32,494-494.
Tangney, 1. P., Stuewig, J., & Mashek, D. 1. (2007). Moral Emotions and Moral
Behavior. Anuual Review ofPsychology(58), 345-372.
Tyran, 1.-R., & Engelmann, D. (2005). To Buy or Not to Buy? An Experimental Study of
Consumer Boycotts in Retail Markets. Economica, 72(285), 1-16.
Webster, F. E. (1975). Determining the Characteristics of the Socially Conscious
Consumer. Journal ofConsumer Research, 2(December), 188-196.