You are on page 1of 19

Why We Boycott: Consumer Motivations for Boycott Participation

Author(s): Jill Gabrielle Klein, N. Craig Smith and Andrew John


Source: Journal of Marketing, Vol. 68, No. 3 (Jul., 2004), pp. 92-109
Published by: American Marketing Association
Stable URL: http://www.jstor.org/stable/30162004 .
Accessed: 27/09/2013 08:34

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Marketing.

http://www.jstor.org

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
JillGabrielleKlein,N. Craig Smith,& AndrewJohn

WhyWeBoycott: Consumer
Motivations
forBoycottParticipation
Althoughboycottsare increasinglyrelevantformanagementdecision making,therehas been littleresearch of an
individualconsumer's motivationto boycott.Drawingon the helpingbehaviorand boycottliterature, the authors
take a cost-benefitapproach to the decision to boycottand presenta conceptualizationof motivationsforboycott
The authorstested theirframework
participation. duringan actual boycottofa multinational firmthatwas prompted
byfactoryclosings. Consumers who viewed the closures as egregious were more likelyto boycottthe firm, though
onlya minority did so. Four factorsare foundto predictboycottparticipation:the desire to make a difference, the
scope forself-enhancement,counterarguments thatinhibitboycotting,and the cost to the boycotterof constrained
consumption.Furthermore, self-enhancementand constrainedconsumptionare significant moderatorsofthe rela-
tionship between the perceived egregiousness of the firm's
actions and boycottparticipation.The authors also
explore the role of perceptionsof others' participationand discuss implicationsformarketers,nongovernmental
organizations,policymakers,and researchers.

Theboycott is thewaywe takeourcauseto thepublic. publicpolicyimplications. Boycotters deliberately use their


Forsurelyifwe cannot findjusticeinthecourts
ofrural "purchase votes" to favor firms with societal
preferred
wewillfindsupport
California, withourbrothers
andsis- consistent with the idea of consumer choice as a
tersthroughoutthenation. impacts,
-Cesar Chavez(qtd.inWhy WeBoycott rationalefor capitalism(Dickinsonand Hollander1991;
[United
FarmWorkers ofAmerica1973]) Smith1990).Yet therehas beenlittleresearchintothefac-
tors thatinfluencea consumer'smotivationto boycott,
We'vetakensignificant actionsto improve thelives, despitetheneedformarketers, boycottorganizers, andpol-
and
opportunities,working conditions of the peoplewho icymakers to betterunderstand these factors.
makeourproduct aroundtheworld,and[we]regularly Friedman(1985, p. 97) definesa consumerboycottas
investinthecommunities wherewe do business. Andwe "an attempt byone ormorepartiesto achievecertainobjec-
do thisso thatconsumers canbuyNikeproducts withthe tivesby urgingindividualconsumersto refrain frommak-
knowledge thattheseproducts havebeenmanufactured
under safeandfairworking conditions. ing selected purchases in the marketplace." The "urging" of
-Nike's response tocriticismin a boycott typically comes from a nongovernmental organi-
NaomiKlein'sNoLogo(Nike2000) zation(NGO) thatis protesting corporatepractices.Thus,
boycottsare an extremecase of a broadercategoryof con-
Boycotts arean intriguing formofconsumer behavior. sumerbehaviorin whichsocial and ethicalissues,suchas
They are unwelcometo marketers yet consistent environmentalism, influence purchasedecisions.Therefore,
withthemarketing concept,becausefirmstargeted a betterunderstanding of boycottparticipation notonlyis
by a well-supported consumerboycotthave apparently usefulin itsown
rightbut also is likely to inform theunder-
failedto sustaina sufficient customerfocus.As a resultof of ethicalinfluences on behavior in
standing buyer general.
greaterpublic attention to corporatesocial responsibility Consumerboycotts dateback at leastas faras thefour-
(CSR) andtheincreasedvulnerability ofbrandsandcorpo- teenthcenturyand have contributed to some spectacular
ratereputations, boycotts have become evermorerelevant successes forrelativelypowerlessgroups.In the United
formanagement decisionmaking.Furthermore, giventhat States,boycottswere the key to unionization(Wolman
theyrepresent a sourceof consumerpowerand a mecha-
1916), and the 1955 Montgomery bus boycottmarksthe
nismforthesocial controlof business,boycottsalso have of themoderncivilrightsmovement
beginning (Friedman
1999). Elsewhere,examplesincludeGandhi'sboycottsof
Britishsalt and clothbeforeIndianindependence and the
JillGabrielle
KleinisAssociateProfessorofMarketing, INSEAD (e-mail: British ofBarclaysBankbeforeitswithdrawal from
N.CraigSmith isAssociate Professor
ofMarketing boycott
jill.klein@insead.edu). South Africa In the the busi-
andEthics, London Business School(e-mail: ncsmith@london.edu). apartheid (Smith 1990). 1990s,
Andrew Johnis ChiefExecutive AJKExecutive
Officer, Consulting ness pressagreedthatboycottswereoftensuccessfuland
(e-mail: andrew.john@aya.yale.edu).The authors thankRichard P. were occurringmore frequently(e.g., The Economist
Bagozzi, PaulaBone,Pierre Chandon, PamScholder Minette
Ellen, E. 1990).1 Recentprominent consumerboycottsincludethe
Drumwright, Randall Heeb,John G.Lynch Jr.,Naufel andthe
Vilcassim,
Behavioral Reading GroupatLondon Business Schoolforhelpfulcom-
ments onprevious drafts
ofthisarticle.
They alsogratefully
acknowledge 'Theincidence ofboycotts andtheir successareinherently hard
thecollaborationoftheir
anonymous corporatesponsor anditsresearch toquantifybecause of difficulties
in identifying callsfor boycotts
agency as wellas financial
andother assistanceprovidedbyLondon by NGOs andtheunderstandable reluctance of firms to report
Business School,INSEAD, andUniversityofNewSouth Wales. salesdeclinesdue toboycotts orto publicize concessions toboy-
cottorganizers.

Journal of Marketing
92 I Journal July2004
ofMarketing, Vol.68 (July
2004),92-109

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
Europeanboycottof Shell because of its plan to sinkthe Callero 1999); and corporatephilanthropy (Piliavin and
BrentSparoil platform at sea andthemulticountry boycott Charng1990).At itscore,thisresearchaimsto understand
of Nikeoverallegedsweatshopconditionsat Asiansuppli- whenand whypeople apparently act againstselfishinter-
ers. As theseexamplessuggest,boycottstodayare more estsforthegood ofothers.
typicallyfocusedon corporatepracticesratherthan on Boycotting is a collectiveact similarto voting,whichis
broadersociopolitical Thisshiftin
goalssuchas civilrights. a prosocialbehaviorin whichtheindividual benefit appears
boycottfocusreflects boththeincreasedpowerofthemod- to be limited;nonetheless, people go to thepolls in large
em transnational and,paradoxically,
corporation theheight- numbers(fora discussionof thecollectiveactionproblem
enedvulnerability ofcorporate reputationandbrandimage, in voting,see, e.g., Blais 2001; Downs 1957). In theearly
and it is consistent
withrecentfindings thata firm'sCSR helpingliterature, helpingtypicallywas not viewed as a
recordaffectsconsumerperceptions of the firm'sbrands collectiveactionproblem,butas the literature has broad-
and products(Brown and Dacin 1997; Sen and Bhat- ened in scope,it has incorporated cases (e.g., voting)that
tacharya2001). It is withsuch a contextin mindthatwe involvecollectiveaction.Likewise,somecharitable contri-
developandtesta conceptualization ofmotivations forboy- butions(e.g.,to local publicgoods,suchas NationalPublic
cottparticipation. Radio in the United States) are examples of collective
action.2Boycotting is also relatedto customer complaining
behavior,thoughcomplaining typically is neither prosocial
Motivations
forBoycott in intentnor collective.In mostcases, a complaintis a
Participation purelyindividualact thatis completely independent of the
Table 1 summarizes priorresearchon consumerboycotts. behaviorof others(Blodgettand Granbois1992; Boote
Most boycottstudieshave been conceptualor descriptive 1998). Nonetheless,the literature identifiesone formof
(case studies),witha focuson boycottorganizersand tar- as
complaint simply exit (i.e., theconsumer decidesto shun
gets ratherthanon the consumer.Only two studieshave thefirm'sproductofferings in thefuture), whichis akinto
reportedempiricalresearchthatfocusesdirectlyon vari- an individualact of boycott(Boote 1998; Hirschmann
ables thatinfluence an individualconsumer'sboycottdeci- 1970). In addition,as withboycotting, thereis a trigger
sion.Kozinetsand Handelman's(1998) netnographic study eventthatpromptsa dissatisfied customerto evaluatethe
suggeststhatboycottparticipation representsa complex relativecostsandbenefits of lodginga complaint(Blodgett
emotionalexpressionof individuality and a vehicle for andGranbois1991; SinghandWilkes1991).
moral self-realization. Sen, Giirhan-Canli,and Morwitz An explanationof helpingthathas receivedextensive
(2001) testa theoreticalframework thatproposesthata fun- empiricalsupport overthepastthreedecadesis thearousal:
damentalquestionunderliesa consumer'sdecisionto boy- cost-reward model(see Dovidioet al. 1991). Accordingto
cott:Will the boycottbe successful?They findthatcon- thisapproach,whena potentialhelperencounters another
sumers'participationdecisions are influencedby their personin distress,thehelperinterprets the seriousnessof
perception of thelikelihoodof theboycott'ssuccess,their thesituation andexperiences arousalbasedon thisinterpre-
susceptibilityto normative social influences(social pres- tation.In response,thehelperassesses thepotentialcosts
sure),andthecostsassociatedwithboycotting. and benefits of helping.The higherthenetbenefitof help-
Sen, Giirhan-Canli, and Morwitz(2001) conceptualize ing(rewardsminuscosts),themorelikelyitis thathelpwill
boycottsas social dilemmas,whereina consumerchooses be given.Ourapproachto boycotting is similar:Consumers
betweentheindividual benefitofconsumption andthewish encounter an initialtrigger event(whichwe referto as a
of a collectiveto refrainfromconsumptionso thatall firm's"egregiousact") thatengendersnegativearousal.In
receivethe sharedbenefitsof a successfulboycott.Simi- response,each consumerevaluatestheexpectedcosts and
larly,a theoreticaleconomicmodelof boycotting by John benefitsofboycotting.
and Klein(2003) treatsboycottparticipation as a collective Table 1 indicatesthatseveralcostsand benefits of boy-
actionproblem,in whichindividualconsumers'incentives cottinghave previously been identified, butmosthave not
to participateare limitedby the knowledgethattheyare beentestedempirically. Thus,drawingon thehelpingliter-
smallrelativeto themarket and bytheiropportunity to free atureand thepriorboycottliterature, we conceptualizethe
rideon theboycotting ofothers. decisiontoparticipate in a boycottas akinto thedecisionto
Consistent withthearticlesby Sen, Giirhan-Canli, and help othersin distress,to contribute to a charity,or to
Morwitz(2001) and Johnand Klein (2003), we viewboy- donateblood. More specifically, we take a cost-benefit
cottingas a formof prosocialbehaviorby which"actions approachto ourinvestigation of boycottparticipation. Fig-
[are]intended to benefitone or morepeopleotherthanone- ure 1 depictsourmodel.
self-behaviorssuch as helping,comforting, sharing,and Perceived Egregiousness
cooperation" (Batson1998,p. 282). Thisis broadlyreferred
to as "helpingbehavior." Overtheyears,a substantial body Consistent
withtheboycottliterature
(e.g.,Friedman1999;
of literaturein social psychologyhas grownfrominitial Garrett
1987; Smith1990; Smithand Cooper-Martin1997),
analysesof emergency helpingto a broadset of studiesof 2Asa technical thecaseofa singlehelper
matter, canoften be
helpingin manydifferent contexts,includingnonemer- reformulatedas a collective
actionprobleminwhichthere
is a set
gencyhelping(Piliavinet al. 1981, 1982); voting,volun- ofpotentialhelperswhodecideto helpwithsomeprobability
teering,andcharitable donations(e.g.,Chambre1987; Pili- (Lynch andCohen1978).In thelanguage ofgametheory,poten-
avin and Charng 1990); blood donations(Piliavin and tialhelpers
might employ a "mixed strategy."

Consumer
Boycott /93
Participation

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
is
ofof target)
of
issue
frame to
social available issue,
specifically:
few
punish and
rejection
preferences)
to More of
over
expectations access or readily
external objectives, frequently
of message names,
and and
and outrage"cost,
justice boycott. patriotism,
consumer
brand to
function product for substitutes low
understanding
a issuesfew "moral laws,
(internal crowd) able
participation,
(as issue, tax
of desire the and time,
self-realization) of
boycott sacrifice, alternatives,
group)
boycotted andidentify, and boycott, with
right of
of from of
influencesthe to at abstention)
Participations efficacy for willing, repeal
likelihood or
motivations anger
issues
social abouteasy(extent change issue
reference selfhood (aware (force
Boycott care to guilt) (connection
(availability
purchase/consumption)
agent success perceived (target social (differentiate(right of
effectiveness)
(preference concerned,
consumer values alternatives
normative from be
smallinstrumental
boycottto boycotting (coming substitutes
cause)
and and of communications) boycotts) consequences (freemust moral
visibility of
(consumers characteristics
consumer
for
Influencing from
boycotting uniqueness (inferior
participation, participationwidespread characteristics
substitutability
and pressure
consensus of of pressure
self-expression characteristics
acceptable)
adverse
Participation
Variables
-.False pro-boycott
overall
-.Expressive pressure exciting/engaging,
substitutes)
-.Susceptibility
-.Perception
-.Free-riding -.Costs"Valence" and
competing
-.No
-.Ease -.Social
-.Seeking
-.Moral
-.Express perceived
"Individuation"
"Cleansing"
Consumers
'Issue
-.Consumer purchased,
sympathy
-.Product Political
materialism)
-.Product -.Guilt
-.Availability
-.Social
-.Sacrifice
1

Boycott as
TABLE
on data
researct well
boycott
boycott
studies) boycotts
sources
boycotts, as
boycotters) with
historical (colonial
Approach U.S.
survey ongoing
Research increase
of active targets movement)
(laboratory secondary
and (netnographic
organizers) (of
from andsources)
interviews research
Prior modeling (price
methods:
reports on
boycott studies

DynamicExperiments
Methodological Multiple1970-1980),
1900-1970),
research (with Ethnographic
(media collection Case
basedsecondary
organizers Historical
nonimportation

behaviorbehavior policy behavior policy behavior

activism activism
Consumer
Orientation Consumer Consumer
and Consumer Consumer
and Consumer

1999) (1998
Klein (1985, (1989)
1995, and
Morwitz (1990)
and GOrhan-Canli,
(2001)
(2003) and 1991, Handelman
John Sen,
Author(s) Friedman Kozinets Smith Witkowski

July2004
ofMarketing,
94 /Journal

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
research

Other

goals
decision.

boycott boycott

with
boycott boycott
powerlessness)
consumer's
the the a reported.
responsibility
ofconsistent of orientation)
Participations
are social
influence
perceived
world"
that investigation
Boycott
awareness leaderawareness
(less
attitudes the
("future
consumers'
success variables
from
boycott of on
values of alienation
values
Influencing
participant participants'
participants' toward
participation inferred
pressure
of directly

Variables -.Cost
-.Social
-.Potential
-.Whether -.Potential
-.Attitudes focus
-.Participant
-.Expectations
-.Participant
-.Credibility -.Participant
reasonably
1 be
(1998)
can

TABLE
Continued
targets (in
boycott) Handelman
variables
versus
secondary and
full start) such
(boycott
Approach
and
reports) (during
of but
supporters
Kozinets
boycott
(media of and
research research
research
boycott
organizers) participation,
(2001),
Methodological
Survey
and
sources SurveySurvey
partial
advance
boycott
Morwitz
and
influence

behavior
behavior that

GOrhan-Canli,
variables
Marketing
managementConsumer
Orientation Consumer on
Sen,

directly
(2003),

Klein
focus
Sturdivani
(1976) not
and
(1987)
and does
John
(1977)
Garrett
Author(s) cited
Miller MahoneyaOnly

Consumer
Boycottparticipation
/95

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
FIGURE 1
Motivatorsof Boycott Decisions

(H6a)

Company Perceived (H1) Boycott(H6b) Brandimage


action egregiousness decision

(H2.) Prevalenceof
(H2b)
(H3b) (H3a) 4 boycotting
(H4b) (H4.)
(H5b) (-15a)

Makea Self-Enhancement (H3) Counterarguments (H4) Constrained


Difference(H2) .Guilt 'Free ride Consumption(H5)
'Boycottefficacy 'Uncomfortable
ifseen 'Smallagent 'Purchasehistory
.Do mypart byothers 'Boycottinducedharm
'Changefirm decision 'Pressureoffamily/friends (e.g.,hurt
jobs)
'Feel better

ourstarting pointis theobservation that,in general,boycott Drawingfromeconomicand psychological


participation.
is
participation promptedby the belief thata firmhas theory,especiallythe cost-rewardmodel of helping,and
engaged in conduct that is strikinglywrongand thathas fromour preliminary studyand theboycottliterature,we
negative and possibly harmful consequencesfor various propose fourdifferent of
categories motivations:
make a
parties(e.g., workers,consumers,societyat large).Typi- difference,
self-enhancement, and
counterarguments, con-
cally,thisperceptionvariesacross consumers:Some will strainedconsumption(Figure1).
considerthefirm'sactionsseriouslywrong,whereasothers
will be less likelyto do so, just as peoplein helpingsitua- Benefits
and Costs
tionsoftenhavedifferent interpretationsofwhether thesce- Make a difference. The motivation to bringaboutsoci-
nariotheywitnessis seriousenoughto precipitate interven- etal changeby participating in a boycottreflects perceived
tion(Lataneand Darley1968; Schwartz1977). To testthis benefitsofboycotting. Boycotters mayhavean instrumental
idea,we conducteda preliminary studyusingmaterials that motivation to changethe targetfirm'sbehaviorand/orto
the
promote long-standing boycott of Nestle over its mar- signal to the firmand othersthe necessityof appropriate
ketingof infantformulain developingcountries. Perceived conduct(Friedman1999; Kozinetsand Handelman1998).
egregiousness differedacross consumers and predicted both Such motivation is typicallytemperedby a general"will-
boycottparticipation and a more negative brand image ingnessto boycott,"influencedby perceivedconsumer
(Klein, Smith,and John2003). Accordingly, we propose effectiveness,thataffects a consumer'sparticipation in any
thatthelevelofperceivedegregiousness has a directimpact specificboycott(Smith 1990). Likewise, Sen, Giirhan-
on boycott participation. Canli,and Morwitz(2001) referto "perceivedefficacy" as
theextentto whicha consumerbelievesthateach boycott
H1:Consumers whofindthefirm's actionstobe moreegre-
participantcan contribute to theachievement of collective
giousaremorelikely toboycott.
goals, and John and Klein (2003) discusshow an exagger-
However,notall consumers whoviewthefirm'sactions ated view of effectiveness mightexplainwhypeople boy-
as egregiouswill participate
in theboycott.In ourprelimi- cottwhenthetarget is unlikelyto notice.All thisresearchis
narystudy,70% of participants ratedtheproblematic com- consistent withfindings thatpeoplearemorecooperative in
pany practice at or above the midpointon a composite social dilemmasiftheyexpectthatthegroupwill attainits
seven-pointscale measureof egregiousness,but only a goals(WienerandDoescher1991).It is also consistent with
minority (45%) said thattheywoulddefinitely or probably researchthatshowsthathelpingis morelikelywhenpoten-
boycott Nestle.
Thus, we ask, do not
Why people participate tialhelpersbelievethemselves competent to helpand have
in boycottsin responseto perceivedegregiousconduct? confidence thattheiractionswillresultinpositiveoutcomes
Fromthisperspective, ourgoal is as muchto explainwhy (e.g.,Midlarsky1984).Thus:
somepeopledo notboycottas itis toexplainwhyothersdo.
H2a:Beliefsinboycotting tomakea differencepredictboycott
Answersto open-endedquestionssuggestedthatboy- participation.Consumers whobelievethatboycotting is
cottersoftenhadmultipleand different motivationsforpar- appropriate andthatitcanbe effectivearemostlikelyto
ticipation,which reflectedperceived costs or of
benefits participateintheboycott.

ofMarketing,
96I Journal July2004

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
As is shownin Figure1. we suggestthatthisand our ture(Dovidioet al. 1991).3Thus,self-enhancement
through
othercost-benefitmotivations affectboycott
directly partic- boycottparticipationincludestheavoidanceof feelingsof
ipation.For example,if a consumerbelievesthatby boy- guiltorthenegativeperceptions of others.
cottinghe or she can changethefirm'sbehavior,thecon- Had:Self-enhancement
factorspredictboycottparticipation.
sumeris morelikelyto boycottbeyondthedirecteffectof The greatertheperceived scopeforself-enhancement
egregiousness.We also proposethatsuchmotivations mod- (andavoidance
ofguiltorsocialcensure),
themorelikely
eratetherelationshipbetweenperceivedegregiousness and is a consumer
toboycott.
theboycott decision:The effects
ofperceivedegregiousness We also expectthatself-esteem moderates therelation-
maybe enhancedor diminished throughinteractions with shipbetween and such
egregiousness boycotting, thatcon-
the cost-benefitmotivations. For make a difference, we sumerswho perceivean opportunity forself-enhancement
expect thatthe between
relationship and
egregiousness boy- are morelikelyto translateperceivedegregiousness into
cottingis strongerforconsumers whobelievethatboycotts boycottparticipation.Althoughto the best of our knowl-
can bringaboutchangethanforconsumerswho do not. edge thisinteractionhas notbeen investigated directlyin
This interactioncan be inferredfromJohnand Klein's thehelpingliterature,
therearehelpingstudiesthatsuggest
(2003) theoreticalboycottanalysisand is consistentwith thatperceivedegregiousness and self-enhancement motiva-
helpingstudiesin whichperceptions oftheseriousness of a tionsinteract.For example,participantswho are made to
situationhavebeen shownto interact withperceivedcom- feelguiltyaremorelikelyto offer helpin an unrelatedsitu-
petenceto help(Crameret al. 1988).Furthermore, research ationto reducetheirnegativefeelings(Carlsmith andGross
on the relationshipbetweenenvironmental concernand 1969), and participants'
experienceof a boostor threatto
relatedprosocialbehaviors,suchas recycling,
has identified self-esteemaffects
perceptionsofegregiousness (McMillen,
a moderatingrole for perceivedconsumereffectiveness Sanders,and Solomon1977). Boycotting whenegregious-
ness is high(equivalently,
givingaid whenand whereit is
(BergerandCorbin1992;Ellen,Wiener,andCobb-Walgren
mostneeded)presumably allowsforthegreatestdegreeof
1991).
self-enhancement.Thus:
H2b:Beliefsinboycotting
tomakea difference
moderate the
between
relationship andtheboycott
egregiousness deci- H3b:Self-enhancementfactorsmoderatethe relationship
sion.Whenthesebeliefsarestrongly betweenegregiousness and theboycott
decision.The
held,therelation-
greaterthescopeforself-enhancement,
thegreateris the
shipbetweenegregiousness and boycotting
is greater
betweenperceivedegregiousness and
thanwhenthebeliefsarelessstrongly
held. relationship
boycotting.
Self-enhancement. AlthoughH2ais consistent withSen, Counterarguments. Althoughthereare benefitsof boy-
Giirhan-Canli, and Morwitz's (2001) focus on the utility cotting, there are also costs.Helpingstudiesshowthatas
gainedfromboycottsuccess,we suggestthatin additionto costsforhelpingincrease,helpingdecreases.For example,
suchinstrumental rewards,thereare also intrinsic benefits Schwartz(1977) assertsthatin theprocessof decidingto
fromboycott participation,
potentially regardless ofboycott help anotherpersonin need,thereis a "defensivestep"of
outcome.There is substantial evidencefromthe helping assessing potentialnegativeoutcomesof helping(e.g.,
behaviorliterature thatpeople's feelinggood aboutthem- injuring orembarrassing thepersonin need).Thus,a poten-
selvesandbeingadmiredbyothersarekeybenefits ofhelp- tialboycotter, evenone whoperceivesthefirm'sactionsas
ing, whereas self-blameand public censureare conse- highlyegregious,mightrefrain fromparticipation if he or
shebelievesthatboycotting couldlead to unintended harm.
quences of not helping(Dovidio et al. 1991). Thus, our
second categoryof cost-benefit motivations Forexample,consumers not
might boycottsweatshopsup-
incorporates
psychosocial variables that are associated with self- pliersbecausetheprotest couldhurtthoseitwas intended to
enhancement: enablesthe boycotter to boost help.
Participation Anothertypeof counterargument
social andpersonalself-esteem eitherby associatingwitha pertainsto the con-
sumer'sperception of whether his or herindividualcontri-
cause or groupofpeopleor simplybyviewinghim-orher-
butionwill playanyrole in achievingthecollectiveaction
selfas a moralperson.Kozinetsand Handelman(1998, p.
goal. In thiscase,thecounterargument to theprob-
pertains
477) observethatboycotting seemsto allow fora bettering thatthe consumer's decisionwill influence
ability boycott
thatis "akinto a hygienically cleansingprocess."Thiscom- the firm'sdecision.Thereare two variations.First,boy-
parison is reminiscent of Smith's (1990) notionthatpoten- cottersmightbelievethattheiractionswill haveno impact
tial boycotters may believe that theyare undera moral because theyare too smallto be noticed(Johnand Klein
to
obligation keep away from the company'sproductsin 2003). This is analogousto theargument thatthereis no
orderto have"cleanhands.- pointin votingbecause any individualvote will almost
Sociallyembeddedexpectations or social pressuresare
also likelyto affecttheguiltor positivefeelingsassociated
withboycotting. The relevanceof social pressureforboy- 3Peoplecanlearnthathelping is a goodanddesirable
behavior
thatis independent
ofsocialapproval (ordirect andthus
reward)
cottparticipation is widelyacknowledged in theboycottlit-
engageinself-reward.
Thishasbeenshown tobe important
tothe
erature(Friedman1999; Garrett1987; Rea 1974; Sen, maintenanceof long-termhelping, suchas repeatedandregular
Giirhan-Canli, andMorwitz2001) and in thehelpinglitera- blooddonation etal. 1982).
(Piliavin

Consumer
Boycott 197
Participation

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
surelynotaffect theoutcomeofan election.Similarly, help- If the sacrificerequiredto help is sufficientlylarge,
ingoftenfailsto occurbecausethepotential helperbelieves increasedseriousnessof the situationwill not necessarily
thathe or she is unableto intervene thereis a
effectively; predicthelping(Dovidioet al. 1991). Similarly,egregious-
sense of a powerlessness to changethe victim'spredica- nessis morelikelyto be relatedto boycotting at lowerlev-
ment.Second,boycotters mightbelievethattheiractions els of constrainedconsumption. We predictthatthisrela-
are unnecessary becausetheycan freerideon theboycott tionship is weaker at high levels of constrained
decisionsofothers.In thehelpingliterature, free-ridingten- consumption because even high-egregiousness consumers
denciesareexaminedin studiesof"diffusion ofresponsibil- mayfindittoo costlyto boycott. Thus,thedegreeto which
ity,"whichfindthattheprobability thata personwill help boycotting constrainsconsumers'consumption influences
someonein needis drastically reducedwhenothersarealso theeffectofegregiousness on theirboycottdecisions:
availableto help (e.g., Latane and Nida 1981). The cost-
H5b:Thedegreetowhichconsumption is constrained
moder-
benefitmodelof helpingsuggeststhatthisoccursbecause atestherelationship
between andtheboy-
egregiousness
thecostsfornothelping(e.g.,guilt,worryaboutthevictim) cottdecision.Thereis a weakerrelationshipbetween
arereducedas a resultoftheexpectation thatthevictimwill egregiousnessandboycottingforconsumers whosuffer
receive help fromothers.Similarly,althoughboycotts thegreatest intheir
constraint consumption.
requirewidespreadparticipation to be effective(i.e., to
reducesales substantially), if theboycottis successfulall Estimated Participation of Others
willreceivethebenefits regardlessof whether theypartici- Sen,Giirhan-Canli, andMorwitz(2001) andJohnandKlein
pated.Thus, some would-be mightfreeride.4
participants (2003) suggest that perceptionsof how manyothersare
about boycotting are negatively boycotting (estimatedparticipation) also affectindividual
H4a:Counterarguments
relatedto boycott The morea consumer
participation. boycottparticipation. Empirical work on social dilemmas
engagesin counterarguments, thelesslikelyis thecon- (e.g.,Wiener and Doescher 1994) suggeststhatan increase
sumer toboycott. in estimated participation likelyto lead to an increasein
is
actualparticipation. In theboycottcontext, thereare many
We proposethatcounterarguments also moderatethe
ways suchan effect can operate. example,morepeople
For
relationshipbetweenperceivedegregiousnessand boy-
takingpartmaygenerateincreasedsocial pressureor may
cotting.Conflictmodels of choice (Ajzen 1996) suggest affectperceptions of boycottefficacy;whenmorepeople
willnotbe reached
thata decision(in thiscase, boycotting)
participate,an individual consumermaybelievethathis or
ifa personfeelspowerless(smallagent),shiftsresponsibil-
her own boycotting is morelikelyto affectthe outcome
ity to others(free rides), or becomes concernedabout believesthattheboycottis close to a
(becausetheconsumer
potentialconsequences.Decision makingcan be impeded
evenwhenthestakes(in thiscase, egregiousness) are high "tippingpoint"at whichthe firmmightwell capitulate).
and can resultin inactionevenwhenthesituation calls for Thus,in additionto a directeffecton boycotting, we might
findthatestimatedparticipation moderatesthe effectof
action(Hogarth1980). Analogously,increasedegregious-
self-enhancement ormakea difference.
ness may fail to translateintoboycotting if counterargu-
Anotherpossibility is thatestimated mod-
participation
mentsloomlargein themindsofconsumers. Thus:
eratesconstrained consumption. When the directcost of
Hob:Counterarguments aboutboycottingmoderatetherela- boycotting is high,people may be particularly averseto
between
tionship andtheboycott
egregiousness decision.
The strongerthecounterarguments,theweakeris the beingexploitedby others'free-riding (Sen, Giirhan-Canli,
and and Morwitz2001; Wienerand Doescher 1991) and thus
relationshipbetweenperceivedegregiousness
boycotting. may be highlyattunedto whetherothersare takingpart.
Finally,higherestimatedparticipation may decreaseboy-
Constrained consumption.The directcostofboycotting cotting because it affects the incentive to freeride.When
a preferred
(i.e., forgoing good) also factorsintothecon- morepeople participate, the boycottis morelikelyto be
sumer'sboycottdecision.Boycotting is likelyto be most successful,and so the temptation to freeride increases
costlyforheavyusersof thetargeted company'sproducts (Johnand Klein2003); thissuggeststhatestimated partici-
beforetheboycott becausesuchconsumers facethegreatest pationcould moderateour free-rider variable.Thus,there
constraint on theirconsumption if theyparticipate in the are theoretical reasonsthatestimatedparticipation might
we
boycott.Thus, anticipate that constrained consumption moderateall fourfactors.However,thelinksare complex;
on boycottparticipation.
has a directeffect thus,our analysisof thisvariableis moreexploratory and
lacksexplicithypotheses.5
H5a.Thedegree towhichconsumption is constrained
predicts
boycott Consumers
participation. whose is
consumption
mostconstrained
byboycottingarelesslikelytoboycott.
5Theseargumentsalsoimply thepossibility
ofthree-wayinter-
actionsbetweenestimated and the
egregiousness,
participation,
thatwe discuss.Wedo notsearchforsuchinteractions
variables
4Although someofthecounterarguments mayappear tobe sim- herebecauseourmodelis already complex.Thisdiscussion
also
plythereverseof themake-a-difference motivation,analyses thepossible
doesnotexhaust as a matter
waysthat, oftheory,esti-
constructs.
revealthattheyaretwodistinct with
Thisis consistent mightaffect
matedparticipation Forexample,
boycotting. con-
findingsinpsychologythatthepositiveandnegativesidesofthe sumersmightinferegregiousness fromestimatedparticipation.
sameattitude (e.g., Cacioppoand Berntson
are oftendistinct Ourdataarenotrichenough amongthesepotential
todistinguish
1994). effects.

July2004
ofMarketing,
98I Journal

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
Brand Image mediacoverage,and some consumersbegan to boycott.?
Two weeksaftertheannouncement, Bremmer's chiefexec-
Consistent withourpreliminary study,we expectthategre-
affectsbrand Consumers whobelievethat utiveofficerwas interviewed in themediaabouttheclos-
giousness image:
a firmhas erredwillhavea morenegativeimageof it than ings.A largedemonstration at a factory closingwas staged
will consumerswho do notjudge its actionsas egregious a weeklater,and a groupof Bremmer employeesand vari-
ous pressuregroups,includingan NGO knownforitscam-
(see Dawar and Pillutla2000; Smithand Cooper-Martin
paignsagainstglobalization, called foran officialboycott.
1997).
Consumerswere asked to boycottall the firm'sproducts,
H6a:Thereis a direct between
relationship and
egregiousness buttwoproductbrandsreceivedthemostmediaattention:
brandimage;thegreater theperceived the
egregiousness, thebrandmadein thefactories to be closed (BrandA) and
morenegative is thebrandimage. Bremmerbrand-nameproductsmade in otherfactories
We also expectthatboycotting damagesbrandimage (BrandB). Bremmer had otherbrands,butonlysomecon-
beyond thedirecteffectsof It is well estab-
egregiousness. sumerswereawareofthis.
lishedin social psychology thatactionscan intensify atti- We hadaccessto sales-tracking dataforBrandA, which
tudesin thedirectionofthebehavior.Bothcognitivedisso- showedan 11% decreasein market share(i.e., 11% ofshare
nance theory(e.g., Festinger1957) and self-perception in thethreepreviousmonths)in thefirsttwoweeksfollow-
theory(e.g., Bem 1972) predictthatundertaking an action ing the announcement of theclosings(beforetheboycott
leads to behavior-consistent
attitudes.
Thus,independent of was "officially"
called).Sales thenrecovered somewhat, but
egregiousnessperceptions, consumers who boycott are forthefourmonthsfollowingthecall to boycott,market
likelyto devalue theirperceptionof the brand,simply sharewas downan averageof approximately 4%. No sig-
becausetheyboycotted. nificanteventsoccurredduringthe data collectionperiod
(thethirdto tenthweeksfollowing thecall to boycott),
and
H6b:Theboycott decisionmediates
therelationship
between
andbrand image. marketshare remainedlower and steady.Sales did not
egregiousness
recoverto near theirpreboycott levels untilfivemonths
Finally,themanagement of targetedfirmsoftencom- aftertheboycottwas announced.
municateswithconsumers to discourageboycottparticipa-
tion.Thus,we also examineresponsesto messagesintended The Study
to countertheboycott. Our studywas conductedby Bremmer'sresearchagency,
whichused questionsthatwe addedto its corporatetrack-
Methods ingsurveys.We didnothavecontrolovertheexactwording
or theformat oftheresponsescales (becauseoftheneedto
The Bremmer Boycott maintainconsistencywith previous Bremmertracking
We testedourhypotheses in an empiricalstudyofan actual, studies).
ongoingboycott. Thiscontrasts withtheworkof bothSen, Subjects.A nationallyrepresentativestratified
random
Giirhan-Canli,and Morwitz(2001), whoconducteda labo- sample of 1216 adultconsumers participatedin the study
ratory study(theyinformed subjectsaboutan ongoingboy- througha telephonesurvey(response rate was 40%).
cottand manipulated factorspredictedto affectparticipa- Respondents wereaskedtoparticipate in "a surveyon some
tion),and Kozinetsand Handelman(1998), who examined companies," and therewas no referenceduringrecruitment
Internet chat-room data (theyselectedsitesforthepreva- to Bremmer orto theboycott.Interviews wereconductedin
lenceof boycottdiscussions).Thus,oursis thefirstquanti- twophases:duringthethirdand tenthweeksaftertheboy-
tative,consumer-focused studyof an ongoingboycottin a cottwas called. Womenconstituted 52.6% of the sample,
naturalsetting.6The advantageof thisapproachis thatwe andtheaverageage was 46.8
capturereal-time reactionsto a boycottin thesocial milieu Measurement. Respondents wereaskedto indicatetheir
in whichit occurred.Althoughwe lack the experimental opinionof thefirmon a three-point scale (from"poor"to
controlof thelaboratory, we gain access to a realboycott,
withall its inherentemotionand publiccontroversy, with-
outneedingto elicitor simulatemoraloutragein an artifi- ?Ourconfidentiality
agreement precludesdirect
identification
of
cial setting. theboycott andthusweprovide
target, nocitationshere(although,
inaccordance withouragreement, wedididentify tothe
thetarget
The boycottin questionwas called againstBremmer reviewersofthisarticle).
Theplantclosures andresultantboycott
(namedisguised),a European-based multinationalfirmthat werehighly events
significant thatresultedinfront-pagenewspa-
sells consumerfood products,primarily throughgrocery perarticlesand majorcoveragein othermedia.The negative
outlets.The announcement oftwofactory closingsoccurred mediacoverageincludedthecompany's plantclosuresbeing
approximately one monthbeforethestartofdatacollection. describedinthepressas "brutal"andthecompany beingdepicted
The closingswere a majoreventthatreceivedextensive as a "symbolofcorporate greed"becauseitwasmaking closures
eventhough itwasprofitable.A government ministerevendecried
theclosuresas "unacceptable."
8Professionals
represented12%ofthesample; mid-tolow-level
6MillerandSturdivant
(1977) studyan actualboycottdueto managers and technicians,15%; clerical
workers, 11%;unskilled
worker buttheyfocuson theeffect
mistreatment, of onefirm's workers,22%;andunemployed orretired,32%.A stratified
sam-
actionson perceptions
of an affiliated
firmanddo notexamine plingapproach was takento ensurethatsampledemographics
consumermotivations
forparticipation. matched thoseofthepopulation.

Consumer
Boycott 199
Participation

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
"verygood").Theywerethenaskedwhether theyhadheard Bremmer"; "I am tempted to boycott,butI don'tknowifI
aboutthefactory closings("Therehavebeenrecentreports will";and"I amnotboycotting theproducts ofBremmer."10
in thepressaboutplannedfactory closingsandjob lossesin Respondents werealso askedto estimatethepercentage of
theBremmergroup.Have you heardaboutthem?").They Bremmer customerswho were boycottingBremmer
werealso askedthreequestionsthatmeasuredtheperceived products.
egregiousness of theclosings.Respondents were askedto We measuredthehypothesized benefitandcostmotiva-
ratetheirconfidence(on a four-point scale of "no confi- on
tors a ten-point scale =
(1 "strongly disagree"and 10 =
denceat all" to "completeconfidence") in themanagersof "stronglyagree").Table 2 liststhe questionsanddescriptive
Bremmer "to not close factoriesexceptwhennecessary" forthemake-a-difference,
statistics self-enhancement, and
and "to ensurethatthe factoryclosingstakeplace in the counterargument items. We measured constrained con-
bestpossibleway fortheworkers." A thirdquestionasked sumptionby previouspurchasefrequency, because heavy
for agreementon a four-pointscale to the statement, purchasers of Bremmer products would pay a higherprice
"Bremmer mustclose certainunprofitable to avoid
factories forboycotting thanwould lightpurchasers.Respondents
putting its entire[product]line in danger."9Respondents wereaskedhow often,in general,theyhad boughtthetwo
were also asked whethertheydisapprovedof Bremmer's focalbrands(A andB), on a four-point scale (1 = "neveror
actions. almostnever,"and4 = "veryoften").
The nextquestionasked about boycottparticipation, Bremmermanagementcommunicatedtwo messages
statingthattherehad been appealsto boycottBremmerin aboutthefactory closings:Threealternative jobs had been
reactionto the factoryclosings.Respondentswere given
threepossibleresponses:"I am boycotting theproductsof
10This wasmodified
question phaseofthesur-
ina subsequent
withtheboycott
9Inthemedia,muchoftheangerassociated first
vey.Interviewers if
wasboycotting;
askediftherespondent
to reports
was attributed was closingfactories
thatBremmer theanswerwas negative, theyaskedwhether he or she was
overall.
beingprofitable
despite temptedtoboycott.

TABLE 2
Descriptive Statistics
Means

Variable Mean (s.d.) Nonboycotters Tempteds Boycotters

Egregiousness 2.77 (.81) 2.60 (.81) 3.00 (.66) 3.23 (.75)

Makea Difference 4.51 (2.67)a 3.54 (2.23) 5.55 (2.36) 7.44 (2.06)
Boycotts meansto makea company
are an effective
changeitsactions. 4.76 (3.14) 3.96 (2.90) 5.48 (2.98) 7.17 (2.84)
Everyoneshouldtakepartintheboycott because
everycontribution, howsmall,is important.
no matter 4.65 (3.21) 3.57 (2.80) 5.91 (2.68) 7.83 (2.68)
Byboycotting,I can helpchangeBremmer's decision. 4.14 (3.29) 3.10 (2.76) 5.20 (3.13) 7.34 (3.03)

Self-Enhancement 2.82 (2.15)a 2.21 (1.76) 3.41 (2.03) 4.75 (2.41)


I wouldfeelguilty ifI boughtBremmer products. 3.27 (3.10) 2.48 (2.63) 4.07 (2.94) 5.76 (3.49)
I wouldfeeluncomfortable ifotherpeoplewhoare
boycottingsaw me purchasing orconsuming
Bremmer products. 2.85 (2.91) 2.52 (2.71) 3.07 (2.75) 4.01 (3.49)
Myfriends/my family are encouraging me to boycott
Bremmer. 2.04 (2.31) 1.76 (1.94) 2.37 (2.58) 2.89 (3.09)
I willfeelbetteraboutmyself Bremmer.
ifI boycott 3.08 (3.01) 2.04 (2.16) 4.14 (2.89) 6.24 (3.54)

Counterarguments 5.67 (2.23)a 6.25 (2.00) 5.10 (2.00) 3.84 (2.24)


I do notneed to boycott
Bremmer; enoughother
peopleare doingso. 4.52 (3.32) 4.83 (3.37) 4.34 (2.95) 3.43 (3.25)
I do notbuyenoughBremmer products foritto be
worthwhile itwouldnoteven be noticed.
boycotting; 5.22 (3.35) 5.53 (3.37) 5.25 (3.07) 3.89 (3.21)
One shouldn't boycottbecause itwillputother
Bremmer jobs indanger. 7.02 (3.05) 7.92 (2.63) 5.89 (2.82) 4.46 (3.09)
I don'tboycottBremmer because itis a (country)
companyand boycotting wouldlead meto buy
products.
foreign 5.89 (3.47) 6.69 (3.29) 4.91 (3.11) 3.59 (3.28)

ConstrainedConsumption(Purchase History) 2.77 (.75)a 2.88 (.66) 2.80 (.68) 2.27 (.95)
BrandA (central) 2.61 (.89) 2.70 (.85) 2.60 (.89) 2.27 (.99)
BrandB (central) 2.93 (.85) 3.06 (.75) 3.00 (.72) 2.28 (1.06)
aDescriptivestatisticsare forthe average ofthe itemswithineach factor.
Notes: s.d. = standarddeviation.

July2004
ofMarketing,
1001Journal

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
offered to each workerin theclosingfactories, and Brem- indicatesthatnotall consumerswho view a firm'sactions
merwas seekingothercompaniesthatwouldbe willingto as wrongparticipate I
in a boycott.'
createjobs at thesitesto be closed. In thesurvey,
respon- A principalcomponents analysis(PCA) withVarimax
dentswereasked whethertheyhad heardthesemessages rotationof the cost-benefit (motivation)variablesfound
("yes"or "no") andwhether theyweremoresympathetic to fourfactors,as we predicted(Table 3). The PCAs con-
thefirmas a resultoftheseinitiatives,on a four-point
scale ductedfactorbyfactorrevealedeach factorto be unidimen-
(1 = "notat all" and4 = "very"). sional.The moststrongly endorseditemswerethecounter-
arguments, particularlytheidea thatboycottingwouldhurt
otherjobs and wouldlead consumersto buyforeignprod-
Results ucts. The least strongly endorsedwere self-enhancement
The Bremmer controversywas wellpublicized:95% of the items,particularlyones associatedwithsocial pressure.As
sample had heardabout thefactory closings.Overall,67% Table 2 shows,all motivations showedtheexpecteddiffer-
of the samplewerenotboycotting, 17% weretemptedto encesacrossboycottgroups(all ps < .001).12
boycott, and 16% were boycotting. Althoughrespondents
overwhelmingly disapproved(81%) of Bremmer'sfactory
closings,mostconsumerswho disapprovedwere nonboy-
cotters(61%). Only 19% of disapprovers were currently
boycotting, and 20% weretemptedto boycottbuthad not 110fparticipants
inthesamplewhowerehighinegregiousness
yetdone so. For mostconsumers, disapprovalof thefirm's (scoredthreeorhigher)andwhodisapproved,25.5%wereboy-
actionsdidnotlead to boycotting. 20.2%weretempted,
cotting, and54.3%werenotboycotting.
We averaged the three measures of egregiousness 12Therewereno substantial
changesovertimeinthemeasures
describedpreviously (Cronbach'sa = .73). The meanegre- reportedin Table2, withtheexceptionthatthelastwaveof
respondentsgavehigher tothebeliefthat"boycotts
ratings arean
giousnessscore was 2.79 (recall thatresponsesto these effective
meanstomakea company changeitsactions"(p < .05),
questionswereon a four-point scale,and thushighernum- andtherewerefewer whoweretempted
respondents to boycott
bersindicategreateregregiousness). Although60% of the becausetheyweremoredecidedon boycott As a
participation.
sampleaveragedthreeor higheron thescale,only22% of we included
precaution, thetimeofinterviewas a controlinour
this groupboycotted(21% were tempted),whichagain regression
analyses.

TABLE 3
PCA withVarimax Rotation: Eigenvalues and Item Loadings

Component
Makea Self- Constrained
Difference Enhancement Counterarguments Consumption

Eigenvalue 3.64 2.03 1.37 1.01


(% ofVariance) (28.01) (15.59) (10.57) (7.78)
Boycotts are an effective meansto makea
companychangeitsactions. 851 .047 026 -.037
Everyone shouldtakepartintheboycott because
everycontribution, no matter howsmall,is important. .753 .224 -.171 -.088
Byboycotting, I can helpchangeBremmer's decision. .762 .361 -.093 -.018
I wouldfeelguilty ifI boughtBremmer products. 286 .724 -.071 -.114
I wouldfeeluncomfortable ifotherpeoplewhoare
boycotting saw me purchasing orconsuming
Bremmer products. 137 .712 .172 -.027
Myfriends/my family are encouraging me to boycott
Bremmer. 007 .756 059 042
I willfeelbetteraboutmyself ifI boycottBremmer. .466 .648 -.129 -.128
I do notneed to boycott Bremmer; enoughother
peopleare doingso. 105 .070 .731 .043
I do notbuyenoughBremmer productsforitto be
worthwhile boycotting; itwouldnoteven be noticed. -.009 .003 631 -.240
One shouldn't boycott because itwillputother
Bremmer jobs indanger. -.310 -.089 632 223
I don'tboycott Bremmer because itis a (country)
companyand boycotting wouldlead me to buy
foreignproducts. -.206 .099 659 264
Purchasehistory BrandA -.030 -.013 049 831
Purchasehistory BrandB -.084 -.099 .065 833
Notes: Numbersin boldface indicatevariables includedin component.

Consumer
Boycott 1101
Participation

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
Regression Analyses boycotted.15 Table 5 also illustratesthe interactioneffect:
We firstindexedouritemsaccordingto ourpredictedfac- Of thisincreaseof 159 (= 339 - 180), we can attribute 50
tors.Make a difference has a Cronbach'sa of .78, and self- (= 230 - 180) to thepartialeffectof the changein con-
enhancement has a Cronbach'sa of .73. We averagedthe strainedconsumption and 81 (= 261 - 180) to thepartial
twopurchasehistory effectof the changein egregiousness. The remaining28
questionsas a measureof constrained
consumption (r = .485, p < .01).13The counterargument (18% of thetotalchange)are attributable to theinteraction
effect.Our ordinalregressionand discriminant analysis
items,whileforming a clearfactorin thePCA, had a Cron-
bach'sa ofonly.61; we return to thisissuesubsequently. approachesyieldedsimilarpredictions andinteractions.
As Table 2 shows,themeanof each of themotivation In Table 4, Models B-E, we examinethecounterargu-
variablesis consistentwith an equal intervaldependent mentsin moredetail.Althoughthefourcounterarguments
variable: In each case, the mean for "tempteds"falls loadedon a singlefactor, theymeasuredifferent objections
betweenthe meansforthe nonboycotting and boycotting to boycotting:free-riding is notthesameas refraining from
or approximately fromthe boycotting because otherjobs will be threatened. Further-
groups,equidistant equidistant
twoextrememeans.(This was thecase acrossbothphases more, Cronbach's a of .61 is less than the generally
ofthedatacollection.)We thusreportsimpleordinary least acceptedcutoff of .70,whichsuggeststhattheitemsshould
linear be
perhaps represented individuallyin theregression equa-
squaresregression (i.e., a probability model) our
in
We also conducted thatdo notrequirean tion.At the same time,the itemsare correlatedwithone
analysis. analyses
interval-scaled discrimi- another(as indicatedby thefactoranalysisand by ther's
dependentvariable-specifically,
nant analysis and ordinal (logit) regression-and we ranging from.18 to .45), so theinclusionofall fouritemsin
obtainedsimilarresults.We centeredpredictorvariables a singleequationpresentscollinearity problems.We thus
ran separateregressions (Models B-E) thatincludedeach
(rawscoreminusmean;CohenandCohen1983).
ModelA in Table4 is ourbenchmark model.It includes counterargument in turn as an individualvariable.The
and all the cost-benefit motivationfactorsas counterargument variables all havesignificantdirecteffects
egregiousness
well as theinteraction termsof each factorwithegregious- on boycotting, and "too small"and "hurtsjobs" showsig-
ness. We includedrespondents' sex in themodel(see the nificantinteractions with egregiousness.All interaction
and added a dummyvariableforthe termsarein thepredicted direction.16
subsequent discussion)
phase of data collection. As H1 predicts, egregiousness has
a directeffecton boycotting, as do thefoursets of cost-
2.77.Themeanvalueforthelowerthird ofthesampleis 1.85.To
benefitfactors.Thereweresignificant interactionsbetween
constructlow egregiousness, we scaleddowntheegregiousness
egregiousness and self-enhancement and between egre- scoreforourentire samplesuchthatthemeanfortheconstructed
giousnessandconstrained consumption. The makea differ- datais 1.85.Thus,wemultiply everyone's scoreonegregiousness
encex egregiousness andthecounterargument x egregious- bya factor equalto approximately 1.85/2.77 = .668 (ouractual
ness interactions are in thepredicteddirection butare not scalingfactor is slightlylower[.663],becausewe truncated the
datasuchthattheminimum scoreis still1).Likewise, toconstruct
significant.
Table5 indicatesthemagnitude of thefindings. We use highegregiousness, we multiply all scoresby a factor of 1.67,
whichensures thatthemeanin ournewconstructed dataequals
ourdatato predicthow manypeople in our samplewould theactualmeanforthetopone-third ofthetruedata.
haveboycotted had egregiousness andourfourfactors been We generated thepredictions usingtheordinary leastsquares
counterfactually loweror higher.14 For example,had our regressionas anindicator function fortherespondents inoursam-
sample had higheregregiousness(such thatmean egre- ple,usingthetruedataforallvariables except thosenoted. Specif-
giousnessfortheentiresampleequaledthetruemeanlevel ically,we usedouroriginal regression to identifya cutoff value
forthetopthirdof thesample),we predictthat261 partici- between boycotting andbeingtempted toboycott. (Wecouldnot
whichis a 45% increasefrom simply sumthepredicted probabilitiesbecauseofthetempteds.)
pantswouldhaveboycotted, Thus,thevaluesin Table5 shouldbe interpreted as follows: If
the original180. Had constrained consumption also been consumers in oursamplehad,counterfactually, hadproportion-
low, we predict that 339 participantswould have atelyhigheregregiousness (suchthatmeanegregiousness was
equaltothatofthetopthird ofthesample), we predict that261
wouldhavebeenabovethecutoff andwouldhaveboycotted. Had
theyalsohada higher scoreon makea difference (suchthatthe
meanwasequaltothatofthetopthird ofthesample), ourmodel
13Ifrespondentsanswered thequestionnaire the
appropriately, predictsthattherewouldhavebeenmorethan500 boycotters
purchase history
questionsshould capture constrained
consump- ratherthan180.
tion.However, thereare twomismeasurement issuesthatcan 15Although theinterpretation of a changein egregiousness is
consumers
arise:First, whoneverpurchased Bremmer's products straightforward (the firm'sactions could have been perceived as
might decidetoreportthemselves as boycotters;
second,respon- moreegregious), changeina motivation
a counterfactual is more
dentsmight thepurchase
havemisinterpreted history as a
question complex. Oneinterpretation is thatthestrength ofthemotivations
questionabouttheir
currentpurchases.As a check,wealsoranour is indeedmutable foreachparticipant. Another interpretation is
analyseswiththeomission ofconsumers whoreported neverpur- thatdifferent valuesof themotivations correspond to different
chasing Bremmer's products; our resultswere essentially subsamples ofthepopulation.
unchanged. 16Although ouranalysis raisesthepossibility ofmisspecifica-
14The low(high)valuesinTable5 scaleeachscoresuchthatthe tionbias,thisappears tobe limited giventhattheresults forthe
meanscoreforthescaledvariable equalsthetruemeanforthe restofthemodelaresimilar inallfivemodels inTable4. Ineffect,
lower(upper)third of theactualsample.Forexample, consider theindividual regressionsprovide anupper boundontheeffect of
In oursample,
egregiousness. themeanvalueofegregiousness is eachindividual counterargument.

102I Journal
ofMarketing,
July2004

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
+ p 0000.1800.02 .00.00 00.450700
.46
=
E:
t 1.35 R2
Factors 4.82
12.797.19
2.34 -4.20
-3.02 -8.741.85
-.75 3.03
Country
Model
Three b .11.10010703 -.10
-.09 -.05
.06.10 Adjusted
-.01

+ a 0000.250003 0001 .00.03 .07.00 .46


=
D:
Jobst 4.81 1.15 2.12 1.79
3.18R2
Factors 12.066.26 -4.12
-2.65 -9.27
-2.16
Model
Hurt
Three b .11.100106.03 .06.11 Adjusted
-.08
-.10 -.06
-.02

+ 0 0000.2200.05 .0001 .00.03 .11.00 .42


=
C:
t
Small R2
1.24 1.97 1.60
Factors 13.286.61
6.11
-2.82 -5.81
-5.43 -2.12
3.25
Model
Too
Three b .14.11010702 -.13
-.08 -.03
-.01 .06.11 Adjusted

A-E
+ D 0000.220003 00000087 .05.00 .42
4 =
B:
Models Ridet - R2
5.86 1.24 2.18 3.22
5.03 5.64 1.97
-
Factors 13.346.78 -.17
TABLE ModelFree
Three b 1311010703 .07 Adjusted
-.12
-.10
-.03
-.00
Regression
D 0000.28000500.280001 .04.00 .47
= .05.
A: <
t 4.62 1.08 R2 p
Factors 1.98 1.07 2.58
12.137.71 2.12
2.96 at
-11.25-4.37
Model
Four
b .1009010702 07.10 Adjusted
-.09
-.01 -.07
-.10 significant
are

that

variables

egregiousness
x indicate

egregiousness
egregiousness
x x
boldface
in
egregiousness
consumption
consumption
egregiousness
egregiousness
x egregiousness
egregiousness
x x x products
x
difference
a ride
ride Numbers
jobs
small jobs
small
Egregiousness
Make
Difference Free
Free
Self-enhancement
Self-enhancement
Counterarguments
Constrained Hurt
Counterarguments
Constrained
Too
Too Hurt Phase
Country
Sex
Country Notes:

Consumer
Boycottparticipation
/109

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
TABLE 5
Illustrationof Magnitude of Effects: Predicted Number of Boycotters

Egregiousness
Low Medium High
Makea difference Low 12 (1%) 52 (5%) 75 (7%)
Medium 110 (10%) 180 (16%) 261 (24%)
High 254 (23%) 387 (35%) 504 (45%)
Self-enhancement Low 52 (5%) 96 (9%) 124 (11%)
Medium 110 (10%) 180 (16%) 261 (24%)
High 190 (17%) 309 (28%) 409 (37%)
Counterarguments Low 177 (16%) 281 (25%) 394 (36%)
Medium 110 (10%) 180 (16%) 261 (24%)
High 74 (7%) 122 (11%) 166 (15%)
Constrained
consumption Low 123 (11%) 230 (21%) 339 (31%)
Medium 110 (10%) 180 (16%) 261 (24%)
High 95 (9%) 154 (14%) 198 (18%)
Notes: The numbersare based on the linearprobability model. In our sample, therewere 180 actual boycottersout of 1108 observations(less
than our totalsample of 1216 because of missingdata). For an explanationof how we calculated the predictions,see Note 14.

Thus,fromModelsA-E in Table4, we findsupportfor TABLE 6


Hl-H5b(supportforH2bwas directional butnotsignificant, Regression Model F Including Estimated
and supportfor Hob was directionalbut not significant, Participation (Percentage of Others Boycotting)
thoughwe foundsignificant differencesforthe counter-
argument variablestoo small and hurtsjobs). We findthe b t p
directeffectsformakea difference
strongest and counter-
Egregiousness .106 4.52 00
arguments.The interactionswere less powerfulbut were Makea difference 089 10.62 00
andall werein thepredicted
oftensignificant, the
direction; x egregiousness
Difference 009 .85 40
strongestmoderators were constrained consumption and Self-enhancement 056 5.27 00
self-enhancement. Self-enhancementx egregiousness 022 1.71 09
Counterarguments -.090 -10.55 00
Estimatedparticipation. On average, respondents Counterargumentsx egregiousness -.008 -.71 .48
believedthat27% ofconsumers wereboycotting, whichis a Constrainedconsumption -.120 -4.89 00
substantialoverestimateof theactualnumberthatreported Constrainedconsumptionx
participation(16%). Boycottersestimatedthat40% were egregiousness -.073 2.36 02
also overestimated Percentageothers 004 4.14 00
participating.Nonboycotters participa- Othersx egregiousness -.001 -.51 61
tion(at 23%). Whenwe includeestimated participation(see Othersx makea difference 001 2.11 04
Table 6), thebasic model (egregiousnessand cost-benefit Othersx self-enhancement 000 .95 34
motivations)remainssubstantiallyunchanged,and the Othersx counterarguments -.001 2.39 02
explanatory powerincreasesonlyslightly. Still,perception Othersx constrainedconsumption 002 1.37 .17
of others'boycottactivityplays a noteworthy role in the Sex 064 1.77 08
model.Estimated has a
participation significant direct effect Phase 087 2.49 01
on boycotting: The greaterthe numberof people a con-
themorelikelywas the AdjustedR2= .49.
sumerbelievedwereparticipating,
at
Notes: Numbersin boldface indicatevariables thatare significant
consumerto join in. Estimatedparticipationalso interacts
p < .05. Mean estimate of the percentage of other people
significantlywith make a differenceand counter- boycotting= 27% (standarddeviation= 19.12).
arguments.17 are consistent
All thesignsof thecoefficients
withourpreviousconjectures abouttheeffects ofestimated
participation. "good" opinionof thefirm,and 2% had a "poor"opinion.
Brandimage.Trackingdataindicatedthatbeforepublic Thispositiveratinghad droppedto 68% by thestartof our
data collection,and thepercentageof respondents witha
knowledgeof the intendedfactoryclosings,Bremmer's roseto 30% the same
negativeopinion during period.Thus,
imagewas extremely positive:96% had a "verygood" or thecontroversy an thathad been
damaged image previously
almostuniversally positive.
17When we includetheargument variables
separatelyin this Higheregregiousnesswas inverselyrelatedto brand
model(as inModelsB-E),theinteractionsbetweenestimatedpar- image (13= -.33, p < .001; we reportstandardized coeffi-
andhurt
ticipation jobs andbetween estimated and
participation cientshereforease of comparison), in supportof H6, (see
country (p < .05),andtheinteractions
are significant withthe
Figure1). Egregiousness theboycott
also predicted decision
other twocounterargument areintheexpected
variables direction
(p = .24fortoosmall,andp = .18forfree-riding,
whichis consis- (13= .31,p < .001). Whenbothegregiousness and theboy-
tentwiththepredictionthattheincentivetofreerideis stronger cottdecisionpredictbrandimage,bothare significant (13=
whenmorepeopleparticipate). -.24, p < .001,and = -.27, p .001,respectively), the
< and

July2004
ofMarketing,
104I Journal

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
directpathfromegregiousness to boycottingdropssignifi- thecompany'sactions.19It was a situationthatmarketers
cantly(from-.33 to -.24, t = 2.19,p < .05). prefer to avoid.
Thus,we findsupportforthepartialmediationof boy- The perceivedegregiousness ofthefirm'sactionswas a
cottingon the relationshipbetweenegregiousnessand powerful predictorofboycottparticipation. The moreegre-
brandimage (Baron and Kenny1986). This findingsug- gious a consumerperceivedthefirm'sbehavior,themore
geststhatboycotting itself(beyondtheeffects of perceived likelytheconsumer was toboycott. However,egregiousness
egregiousness)predicts brandimage,in support ofH6b.The notwithstanding, mostof thesamplewas notparticipating
effectof boycotting on brandimage remainssignificant in theboycott. To helpunderstand this,we drewon theories
evenifall directeffects andinteractions(fromModelF) are of prosocialbehaviorand the boycottliterature to testa
includedin theregression =
03 -.21, p < .001).18Froma cost-benefit approach boycottparticipation. found
to We
different
angle,amongparticipants who werehighin egre- thatin additionto perceivedegregiousness, consumers need
giousness(scores of three or greateron thefour-point
egre- to believethatboycotting is an appropriate and effective
giousnessscale), 72% of boycottersassigneda "poor"rat- response(makea difference). Consumers also needtorealize
ingto Bremmer, whereasonly24.2% ofnonboycotters gave intrinsicrewardsofboycottparticipation, potentiallyboost-
thisrating.These figuresare muchhigherthanthe pre- or
ing maintaining self-esteem by, for example,avoiding
boycott,fullsample"poor"ratingof 2%. guiltand responding to socialpressure(self-enhancement).
This has a directeffectand moderatestheeffectof egre-
Other Results
giousnesson boycott participation. Furthermore, consumers
Corporatemessagesand egregiousness.Respondents also takeintoaccountthecost of a preferred productfor-
whohadheardthatBremmer hadoffered jobs to
alternative gone (constrained consumption) and the costs of boycott-
dismissedworkersgave significantly loweregregiousness inducedharmsand doubtsaboutwhetherparticipation is
ratingsthandid respondents who had notheardthiscom- necessary(counterarguments). Many of these costs also
munication (m = 2.58 and2.86,respectively, 412031= 5.04, moderate theeffectofegregiousness on boycotting.
p < .001). We foundsimilarresultsforthe messagethat Consumers'estimatesof supportfor a boycottalso
Bremmer had triedto findotherfirmsto employthework- appearto influenceparticipation directly, butthiseffectis
ers (m = 2.65 and 2.86, respectively, t[1204] = 4.29, p < moderated bycounterarguments (peoplewhocounterargued
.001). For respondents who had heardeithermessage,a wereless likelyto boycottwhentheybelievedthatmany
positivereactionto the messagewas inverselyrelatedto otherswereboycotting) and by makea difference (people
egregiousness (r = -.32, p < .001 andr = -.19, p < .001 for who werehighin make a difference weremorelikelyto
otherjobs and otherfirms, respectively).However,respon- boycottiftheybelievedthatmanyotherswereboycotting).
sivenessto corporatemessagesdid notmoderatetherela- The (notsignificant) directional finding fortheinteraction
tionshipbetweenegregiousness and the boycottdecision betweenestimated participation and constrained consump-
(interaction termswere not significant). Thus, corporate tion supportsSen, Giirhan-Canli, and Morwitz's(2001)
messageswereeffective in reducing egregiousness,butthey contention thatwhenboycotting costsare high,peopleare
did notdisrupttherelationship betweenegregiousness and particularly averseto beingexploitedby others'free-riding
boycotting. and are less likelyto boycott.Furthermore, theremaybe
Demographics.Age was not a predictor of egregious- individualdifferences withregardto boycottpropensity,
ness or boycotting (r = -.04, notsignificant,and r = -.06, includingthatwomenare morelikelyto boycottthanmen.
not significant, respectively). sex
Respondents' predicted We also foundthatthe act of boycotting harmedbrand
boycotting (x2 = 11.01,p < .01); 19% of womenboycotted image beyond the effects of perceivedegregiousness.
comparedwith 13% of men.Furthermore, in the models Finally,althoughfirmcommunications did not lessen the
presented in Tables4 and6, sex was significant. An exami- effectof egregiousnesson boycotting, theyappearedto
nationofmeandifferences betweenmenandwomenshows lowerlevelsofperceivedegregiousness.
thatwomentendto be stronger on themotivations forboy-
Managerial Implications
cottingand thatthereare significant differences in per-
ceivedegregiousness, makea difference, self-enhancement, At firstglance,ourfindings mightappearto be encouraging
andfree-riding. formanagersof targeted firmsand disappointing forman-
agersof NGOs thatcall forboycotts,giventheboycott's
highprofileand the relativelylow level of participation
Discussion (16%). Keepingconsumersfrombecomingboycotters is a
Our studyinvestigated a high-profilesocial-issueboycott. keyconsideration forfirms. Askingwhyconsumersdo not
The boycotttargethas been widelycondemnedand has boycottis likewisecriticalforNGOs. Our findings suggest
receivedextensivenegativecoveragein nationalprintand thatevenhighlevelsofperceivedegregiousness areinsuffi-
broadcastmedia.Respondents werewell awareoftheissue cientto motivate becausecost-benefit
boycottparticipation
leadingto theboycottand overwhelmingly disapproved of motivations forparticipation
playa keyrole.

19Asfurther evidence
(indirect) of perceived a
egregiousness,
18Analternative
model,in whichbrandimagemediatedthe national
pollat thetimeoftheboycottfoundthatnearly
nineof
between
relationship andtheboycott
egregiousness was
decision, tenpeoplejudgedit "unacceptable"
forprofitable
companiesto
notsupported. makeemployees redundant.

Consumer
Boycott /105
Participation

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
On closerexamination, however, theboycottappearsto (Smith1990).Ourresultssuggestthatthisis a highlyineffi-
haveinflicted damageon Bremmer. First,theboycott repre- cientstrategy becauseNestle'sinitialnonresponse enraged
sentedsubstantial lostsales in a highlycompetitive market activistsandprobably harmedbrandimageforthelongrun,
(retailauditevidenceof sales declinesduringtheboycott whereasits eventualbackdownprobablyincreasedper-
corresponded to the level of participation reportedin the ceived efficacy. However,the company'sestablishing the
study);Table 5 revealsthattheloss couldhavebeenmuch NestleInfantFormulaAuditCommission, an impartial out-
worse.Second,theboycott providedshopperswitha reason sidegroupofsocialauditors, was widelycreditedwithless-
to trycompetitors' products.Third,participants' attitudes eningsupportforthe boycott,in partby reducingwide-
wereprobablyhardenedagainstBremmerthrough boycott spreadperceptions of egregiousconductby Nestle(Pagan
participation. Perhapsthe most consequentialfindingfor 1986). Nonetheless, a firm'searlyand easy capitulation to
thefirmis thatbrandimagewas harmedamongbothnon- NGO demandsmightsignalweaknessand enhanceboy-
boycotters and boycotters. This indirecteffectmaybe far cotters'sensethattheycan influencethefirmand makea
moresignificant in thelongrunthanthedirectloss ofsales. difference. Thiscouldpromptfurther demandsas theNGO
Managersof targeted firmsmustpay attention to nonboy- ups the ante. For example,afterbeing threatened witha
cottersas wellas boycotters. global boycott, Starbucks quickly agreed to meet the
More fundamentally, thereis empiricalsupportforthe demandby GlobalExchangethat1% of itscoffeesales be
argument boycotts serveas a mechanism
that can by which certifiedas "fairtrade";GlobalExchangeis nowseekingto
consumerscan hold firmsaccountableforperceivedCSR increasethatpercentage to 5% (Sullivan2003).
failings. As wellas beingofpracticalsignificance toNGOs, Good customerrelationships facilitatecommunication,
thisfinding is of theoretical significance because it demon- andpriorinvestments in branding anda sociallyresponsible
stratesenhanced consumersovereignty(Smith 1990). image can serve as a form of "insurance" to counteract
Changesin companypolicyfollowing boycotts ofShelland information about egregiousconduct(Dawar and Pillutla
Nike,forexample,areconsistent withtheideaoffirms being 2000; Klein and Dawar 2003). Communications shouldbe
heldaccountable inconsumer markets andwiththispressure directed atbothnonboycotters andboycotters toreduceper-
translating intoa "businesscase" forCSR (Smith2003). ceptionsof egregiousnessand therebyto protectbrand
We proposea variationof the well-known awareness, image and to lowerboycottparticipation. For theNGO's
trial,repurchase modelof buyerbehavioras a framework part,it mustbe readyto respond.This leads to our first
for boycottmanagement.Our awareness,egregiousness, propositionsforfirms(Pif)and NGOs (P1n),whichare
boycott(AEB) modelincorporates ourfindings on motiva- shownin Figure2.
tionsforboycottparticipation and offerspropositions for Thenextstepis toexaminewhether perceived egregious-
managersof both firmsand NGOs. Althoughwe offer nesstranslates intoboycott participation.Here,ourfindings
propositions to each partyseparately, each shouldbe aware on costsand benefits come intoplay.If consumersbelieve
of theotherparty'spropositions as a sourceof insighton thattheboycott willinfluence thefirm, theyaremorelikely
opposingstrategies and tactics.Figure2 depictsthemodel to participate(makea difference). As P2fsuggests(Figure
withdatafromtheBremmer boycott. 2), firmscan communicate thattheboycottis unlikelyto be
The firststepforthefirmis themeasurement of aware- successful.This messagemustbe communicated carefully,
ness. For many boycotts,awareness is extremelylow forshowingthatconsumerobjectionshavebeenheardis a
(Friedman1999).Firmsmightnonetheless chooseto treata keycomponent of P2f.Corresponding strategies forNGOs
boycottcall as a warningsignal and investigatethe are suggestedin Pen.Consistentwithbuildinga sense of
allegedlyegregiousconduct.Even a smallnumberof pro- consumerefficacyand power, the Ethical Consumer
testorsmighthave a validcriticism of thefirm'sbehavior. ResearchAssociationlistsrecentboycottsuccesseson its
However,if awarenesslevelsare highor growing, thenext Web site.Anothereffective strategy forNGOs is probably
step is research to measure the levelof perceivedegregious- to startsmalland buildon successesmoreeasilyachieved.
ness and to gain a betterunderstanding of whyconsumers University apparelwas an initialfocusof campaignsin the
findthefirm'sactionsobjectionable. The firmcan thenfor- UnitedStatesagainstsweatshoplabor,and severaluniver-
mulateits boycottstrategy and decide whetherto change sityadministrations quicklyaccededto student demandsin
current practice,engagein mitigating actions,or communi- thefaceofcampusdemonstrations (Brixey2000).
cate thereasonsbehindits actions(Smith1990). Bremmer Our findingssuggestthatpeople boycottto feel good
was unwilling to reconsider itsfactory-closure decision,but about themselves(self-enhancement), but theyare less
manyfirmsarenotso committed to a policythatprompts a likelyto participateiftheybelievethattheboycotthas neg-
boycott, andthisis likelytohavea significant impacton the ative outcomes(counterarguments). Althoughmanagers
firm'sstrategy and on boycottsuccess(Garrett1987). The should refrain from directly challenging the self-
increasing powerof NGOs emphasizestheneedfora firm enhancement valueofsocialaction,theircommunication of
tobe particularly carefulin assessingitspolicycommitment harmfuleffectsof boycottingmay lessen the feel-good
(Spar and La Mure 2003). aspectsof participation and reducethe linkbetweenper-
Ourfindings illuminate thefirm'skeystrategy decision ceivedegregiousness and boycotting. For example,a firm
of whether to fightor to acquiesceto NGO demands.For boycotted because of its low wages (byWesternstandards)
example,in theNestleboycott, thecompanyinitially chose to overseasworkers may counter thata "successful" boycott
to ignoreits critics,then foughtthem,and eventually would resultin the closure of factories,leaving former
acceded to theirdemandsin an embarrassing climbdown employeesin poverty. Thus,we proposeP3f(Figure2).

106I Journal July2004


ofMarketing,

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
FIGURE 2
AEB Model

Awarenessof Perceived Boycott


ContentiousAct Egregiousness/Disapproval Participation

96% 81% 16%

P1f:Firmswithhigh/growing awarenessofcontentious P2f:Whenfirms act is


are certainthatthecontentious
acts shouldengage incommunicationsthatreduce theyshouldcommunicate
unavoidable, thatthey
perceivedegregiousnessbycommunicating have heardconsumerconcerns,havechosenthe
changesinthefirm's practices, a
byproviding bestcourseofaction,and willnotchangetheir
convincing for
explanation existing or by
practices, decision.
actions.
ameliorating
publicizing
Pen:NGOs shouldincreaseconsumers' feelingsof
P1n:AnNGO promoting a boycott mustbe readyto efficacyand power.Theyshouldemphasizethe
counterthetargetfirm's the
claimsjustifying importance ofhighparticipationlevelsand the
egregiousact and to respondto anyameliorating beliefthattheboycottwillattainitsgoals ifpeople
actions.Furthermore, negativeconsequencesof Citingpastcases ofsuccessfulboycotts
participate.
thefirm'sactionsshouldbe stressedto enhance can enhanceconsumers' faiththattheiractionswill
perceptionsofegregiousness. producedesiredchange.

P3f:Firmsshouldconveynegativerepercussions
ofthe
boycottto consumers.

Pan:NGOs shouldstriveto convinceconsumersthat


partieswillbe betteroffiftheboycott
affected is
successfuland shouldpreparefact-basedrebuttals
to companyarguments aboutboycott-induced
harms.
firms
P4f:Boycotted shouldcontinueto promote
positive
aspectsoftheirproducts
to increasetheperceived
costofboycott
participation.
Pon:NGOs shouldpublicizetheease ofparticipation,
theavailability
ofcomparablesubstitutes,
and the
smallsacrifice to bringaboutan ostensibly
required
greatergood.

Notes: Percentage values referto the Bremmerboycott.

As forNGOs, theyshouldpreparerebuttals to company ney2001). Moregenerally, NGOs shouldseekmediaatten-


arguments aboutboycott-induced harms(P3nin Figure2). tionto theboycottbecausethiscan be expectedto increase
In responseto theclaimthatblackswerehurtmostbyboy- bothawarenessand socialpressureforparticipation.
cottsoffirmsassociatedwithapartheid SouthAfrica,NGOs Boycottingentailsa sacrificeby the consumer(con-
responded that the goal of freedom from apartheidwas strained consumption).ForNGOs, thetaskis tojustifyand
moreimportant (Smith1990). Furthermore, a key insight minimizethissacrifice.Firmsoftencutadvertising duringa
foran NGO thatcalls fora boycottis thatits motivations boycott (as Bremmer did) becausetheydo not wish to draw
neednotbe identicalto thoseof individualconsumers. An attention tothebrandorbecausetheybelievethattheadver-
NGO witha strongly instrumental
goal may best achieveits tisingspendwouldbe wasted.The oppositestrategy maybe
aimsbyencouraging self-enhancementthrough boycotting. warranted. Well-crafted could
advertising strengthen brand
Thus,withsubtlety, the NGO could implythatfailureto commitment and reinforcepositive associationsto the
wouldlead to feelingsof guilt.Equally,theuse
participate brandin memory, thusincreasingthecostofboycotting and
ofcelebrityendorsements byNGOs can be expectedto help helpingcounteractthe effectsof negativeinformation
boostconsumers' self-esteemthroughboycottparticipation. (Ahluwalia,Burnkrant, and Unnava2000; Tybout,Calder,
For example,Bianca Jaggerhas lent her supportto the and Sternthal 1981).Furthermore,advertisingthattapsinto
Exxonboycott calledoveritsenvironmental policies(Whit- values such as health and safetymightbe beneficial,

Consumer
Boycott 1107
Participation

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
becauseforgoing a productwiththeseattributes
mayreduce It is unlikelythatwe haveuncoveredall potential moti-
the possible self-enhancement of boycottparticipation. vationsforboycott participation,and further researchcould
Thus,we propose134f andPon(Figure2). investigate othercosts and rewardsof boycotting thatwe
mayhavefailedto capturein ourmodel.We didnotinvesti-
Study Limitationsand Directions forFurther gate theboycotter's decision-making process,and particu-
Research
larlywhether egregiousness leads to a firm'sproductbeing
Ourresearchfacessomelimitations thatstemfromtheexi- excludedfromtheconsumer'sconsideration setor whether
gencies of researchingan actual boycottin progress. thefirm'segregiousconductis tradedoffagainstproduct
Althoughour measureswereinformed by ourpriorstudy, attributes. The answermay lie in the level of perceived
we had neithertheluxuryof pilotingsome measuresnor egregiousness: Perhapsat moderatelevelsofegregiousness,
absolutediscretion oversurveycontent, administration,and theconsumertradesoffthefirm'sconduct,butat highlev-
theprecisewordingof questions.However,we believethat els,theconsumer excludestheproductfromconsideration.
theselimitations do notmaterially influenceourresultsand, The studyalso has implications forresearchon corpo-
on balance,are minorcomparedwiththe opportunity to rateassociationsand CSR-relatedissues.Ourfindings with
a
study real,ongoingboycott. to
respect boycotts may well extend to thebroader category
Thereareissuesofgeneralizability. Ourstudyis ofa par- of ethicalinfluences on consumerbehavior,whichimplies
ticularboycotton a specificissue.The egregiousact in this scopeforexploring theroleofa similarsetofmoderators in
case (i.e., factoryclosings)maybe associatedwithgreater themodelsof,forexample,Brownand Dacin (1997) and
self-interest thanis thecase forboycottsprompted by,for Sen and Bhattacharya (2001). Our workmightalso con-
example, animal The was
rights. boycott particularly highpro- tribute toresearchon complaining behavior, thussuggesting
file(whichis all themoreusefulto theextentthatwe were motivationsthat may underliecomplaining(e.g., self-
interested in explainingwhypeopledo notboycott), butitis enhancement). Furthermore, our framework could be used
possible, or even probable,thatour resultsarenotrelevant to to studyothercollectiveactionsituations, suchas voting.
themanycalls to boycottthatconsumers largelyignore and Finally,in lightof ourclaimson self-enhancement, we
thatmanagement canalso safelyignore.In addition,
although believethatitis appropriate to identifyboycottingas partof
ourdataarefroma nationally representativesample,theyare a broaderformof "symbolicnonconsumption." Writing on
froma specificcountry. We doubtthatthetrigger forthis possessionsandsymbolicconsumption, Belk (1988,p. 139)
boycottwouldhave had thesame resonancein theUnited states,"Thatwe are whatwe have ... is perhapsthemost
States,forexample.We have soundtheoretical reasonsto basic and powerful factof consumerbehavior." Whatdoes
believethatthemotivations we uncovered arequitegeneral, it meanwhenconsumers choosenotto consumea product
butthisremainsto be confirmed byresearchon otherboy- becauseofa socialissueassociatedwiththeproducer? Per-
cottsin different countries,prompted bydifferentissues. hapswe boycottbecausewe arealso whatwe do nothave.

REFERENCES
Ahluwalia, Rohini,RobertE. Burnkrant, and H. Rao Unnava Behavior,"Journal of Consumer Satisfaction,Dissatisfaction
(2000),"Consumer Response toNegative Publicity:TheMod- andComplaining Behavior,5, 93-103.
erating RoleofCommitment," Journal ofMarketing Research, Boote,Jonathan(1998),"Towards a Comprehensive Taxonomy
37 (May),203-214. andModelof Consumer Complaining Behavior," Journalof
Ajzen,Icek(1996),"TheSocialPsychology ofDecisionMaking," ConsumerSatisfaction, Dissatisfaction and Complaining
in SocialPsychology: Handbook ofBasicPrinciples, E. Tory Behavior,11,140-51.
HiggensandArieW. Kruglanski, eds. NewYork:Guilford Elizabeth
Brixey, (2000),"Sweatshop Watchdog toMapItsStrat-
Press,297-325. egyWhentheNewWRCHoldsItsFirstMeeting inNewYork
Baron,ReubenM. andDavidA. Kenny(1986),"TheModerator- ThisWeekend," WisconsinStateJournal, (April6),Al.
MediatorVariableDistinctionin Social Psychological Brown,TomJ.andPeter A. Dacin(1997),"TheCompany andthe
Research:Conceptual, Strategic,and StatisticalConsidera- Product:CorporateAssociationsand ConsumerProduct
tions," JournalofPersonality andSocialPsychology, 51 (6), Journal 61 (January),68-84.
1173-82. Responses," ofMarketing,
and ProsocialBehavior," in Cacioppo,JohnT. andGaryG. Berntson (1994),"Relationship
Batson,Daniel(1998), "Altruism
Handbookof Social Psychology, Vol. 2, DanielT. Gilbert,
Between AttitudesandEvaluative Space:A Critical Review,
withEmphasis on theSeparability of PositiveandNegative
SusanT. Fiske,andGardner Lindzey, eds.Boston:McGraw-
Substrates,"
Psychological Bulletin,115(3),401-423.
Hill,282-316. J.MerrillandAlanE. Gross(1969),"SomeEffects of
Belk,RussellW. (1998),"Possessions andtheExtended Self," Carlsmith,
GuiltonCompliance," Journal ofPersonality andSocialPsy-
Journal ofConsumer Research,15(September), 139-68.
11(3),232-39.
Bem,D.J. (1972), "Self-Perception Theory," in Advancesin chology,
SocialPsychology,Vol.6,L. Berkowitz,ed.New Chambre,S.M. (1987),GoodDeedsinOldAge:Volunteering by
Experimental theNewLeisureClass.Lexington, MA:Lexington Books.
York:Academic Press,1-62.
Berger,Ida E. andRuthM. Corbin(1992),"Perceived Consumer Cohen,JacobandPatricia Cohen(1983),Applied MultipleRegres-
Effectiveness andFaithin Others as Moderators ofEnviron- sion/Correlation
Analyses fortheBehavioral Sciences.Hills-
mentally Responsible Behaviors,"Journal ofPublicPolicy& dale,NJ:Lawrence Erlbaum Associates.
Marketing, 11(Fall),79-89. Cramer,R.E., M.R. McMaster, RA. Bartell,and M. Dragna
Blais,A. (2001),ToVoteorNottoVote:TheMerits andLimits of (1988), "SubjectCompetenceand Minimization of the
RationalChoiceTheory. Pittsburgh,PA: Universityof Pitts- BystanderEffect,"Journal ofAppliedSocialPsychology, 18
burgPress. (13),1133-48.
Blodgett,JeffreyG. andDonaldH. Granbois (1992),"Towards an Dawar,NirajandMadanM. Pillutla (2000),"Impact ofProduct-
Integrated ConceptualModel of ConsumerComplaining HarmCriseson BrandEquity:TheModerating RoleofCon-

July2004
ofMarketing,
1081Journal

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions
sumerExpectations," Journalof Marketing Research,37 Miller,Kennethand Fredrick Sturdivant (1977), -Consumer
(May),215-26. Responses to SociallyQuestionable Corporate Behavior: An
Dickinson, RogerandStanley C. Hollander (1991),"Consumer EmpiricalTest," Journal ofConsumer Research, 4 (June), 1-7.
Votes," Journal ofBusiness Research, 22 (4), 335-46. Nike(2000),"Nike'sResponseto No Logo(byNaomiKlein),"
Dovidio,John F.,JaneA. Piliavin, SamuelL. Gaertner, DavidA. (March8, 2000),[availableat http://www.nikebiz.com/labor/
Schroeder, and RussellD. ClarkIII (1991),"TheArousal: nologo_let.shtml].
Cost-Reward ModelandtheProcess ofIntervention: A Review Pagan,RafaelD. (1986),"TheNestleBoycott: Implications for
oftheEvidence," inProsocial Behavior, Margaret S. Clark,ed. StrategicBusinessPlanning," Journal ofBusinessStrategy, 6
Newbury Park,CA: SagePublications, 86-118. (Spring),12-18.
Downs, A. (1957),AnEconomic Theory ofDemocracy. NewYork: Piliavian,JaneAllynandPeterL. Callero(1999),Giving Blood:
Harper andRow. TheDevelopment ofan Altruistic Identity. Baltimore: Johns
TheEconomist (1990),"Boycotting Corporate America," (May Hopkins University Press.
26),69-70. andHong-Wen Charng (1990),"Altruism: A Reviewof
Ellen,Pam Scholder, JoshuaLyleWiener,and CathyCobb- Recent Theory andResearch," Annual Review ofSociology, 16,
Walgren (1991),"TheRoleofPerceived Consumer Effective- 27-65.
nessin Motivating Environmentally ConsciousBehaviors," -, John.F. Dovidio,S.L. Gaertner, and R.D. ClarkIII
Journal ofPublicPolicy& Marketing, 10(Fall),102-117. (1981),Emergency Intervention. NewYork:Academic Press.
Festinger, L.A. (1957),A Theory ofCognitive Dissonance. Stan- -, and -, (1982), "Responsive
ford, CA: Stanford University Press. Bystanders: TheProcessofIntervention," in Cooperation and
Friedman, Monroe(1985),"Consumer Boycotts in theUnited Helping Behavior: Theories andResearch, V.J.DerlegaandJ.
States,1970-1980:Contemporary Eventsin Historical Per- Grzelak,eds.NewYork:Academic Press,279-304.
spective," Journal ofConsumer Affairs,19(1),96-117. Rea,SamuelA.,Jr.(1974),"TheEconomics ofa Consumer Boy-
(1991),"Consumer Boycotts: A Conceptual Framework cott,"Journal ofEconomics andBusiness, 27 (1), 89-92.
and ResearchAgenda,"Journalof Social Issues,47 (1), Schwartz, ShalomH. (1977),"Normative Influences onAltruism,"
149-68. in Advances in Experimental SocialPsychology, Vol. 10,L.
(1995),"American Consumer Boycotts in Responseto Berkowitz, ed.NewYork:Academic Press,221-79.
RisingFood Prices:Housewives' Protests at theGrassroots Sen,SankarandC.B. Bhattacharya (2001),"Does DoingGood
Level,"Journal ofConsumer Policy,18(1),55-72. AlwaysLeadtoDoingBetter? Consumer Reactions toCorpo-
(1999),Consumer Boycotts.NewYork:Routledge. rateSocialResponsibility," Journal ofMarketing Research, 38
Garrett,DennisE. (1987),"TheEffectiveness ofMarketing Policy (May),225-43.
Boycotts: Environmental Opposition toMarketing," Journal of , Zeynep Giirhan-Canli, andVickiMorwitz (2001),"With-
Marketing, 51 (April),46-57. holding Consumption: A SocialDilemma Perspective onCon-
Hirschman, Albert 0. (1970),Exit,Voice, andLoyalty: Responses sumer Boycotts," Journal ofConsumer Research, 28 (Decem-
to Declinein Firms,Organizations, and States.Cambridge, ber),399-417.
MA:Harvard UniversityPress. Singh, JagdipandRobert E. Wilkes (1991),"ATheoretical Frame-
Hogarth, RobinM. (1980),Judgment andChoice.NewYork:John workforModeling Consumers' Response toMarketplace Dis-
Wiley& Sons. satisfaction,"Journal ofConsumer Satisfaction, Dissatisfaction
John,Andrew andJillG. Klein(2003),"TheBoycott Puzzle:Con- andComplaining Behavior, 4,1-12.
sumerMotivations forPurchase Sacrifice,"Management Sci- Smith, N. Craig(1990),Morality andtheMarket: Consumer Pres-
ence,49 (9), 1196-1209. sureforCorporate Accountability. London: Routledge.
Klein,JillG. andNirajDawar(2003),"TheRole of Corporate (2003),"Corporate Social Responsibility: Whether or
SocialResponsibility in Consumers' Attributions andBrand How?"California Management Review, 45 (Summer), 52-76.
Evaluations After a Product-Harm Crisis,"paperpresented at andElizabeth Cooper-Martin (1997),"EthicsandTarget
theAssociation forConsumer Research AnnualConference, Marketing: TheRoleofProduct HarmandConsumer Vulnera-
Toronto (September). Journal
bility," ofMarketing, 61 (July), 1-20.
, N. CraigSmith, andAndrew John(2003),"Exploring Spar,DeboraL. andLane T. La Mure(2003),"ThePowerof
Motivations forParticipation in a Consumer Boycott," in Activism:Assessing theImpact ofNGOsonGlobalBusiness,"
Advances in Consumer Research, Vol.29, SusanBroniarczyk CaliforniaManagement Review, 45 (Spring), 78-101.
and KentNakamoto, eds. Provo,UT: Association forCon- Sullivan,James (2003),"CallitStarbucking, theFineArtofHat-
sumer Research, 363-69. ingYourLocalOutlet oftheSeattleCoffeehouse Chain:Why
Kozinets, Robert V. andJayM. Handelman (1998),"Ensouling Is Starbucks theBrandWe Love to Hate?"San Francisco
Consumption: A Netnographic Exploration of Boycotting Chronicle, (August 13),Dl.
Behavior," inAdvances inConsumer Research, Vol.25,J.Alba Tybout, AliceM., BobbyJ.Calder,andBrianSternthal (1981),
andW.Hutchinson, eds.Provo, UT:Association forConsumer "UsingInformation Processing Theoryto DesignMarketing
Research, 475-80. Strategies," Journal of Marketing Research,18 (February),
Latane,BibbandJohnM. Darley(1968),"GroupInhibition of 73-80.
Bystander Intervention," Journal of Personality and Social United FarmWorkers ofAmerica (1973),Why WeBoycott. Keene,
Psychology, 10(3), 215-21. CA: United FarmWorkers, AFL-CIO.
andSteveNida(1981),"TenYearsofResearch onGroup Whitney, Anna(2001),"Jaggers to JoinPicketof Esso Stations
SizeandHelping," Psychological Bulletin,89 (2),308-324. overKyotoLink,"TheIndependent, (May 8), [availableat
Lynch, John G.,Jr., andJerry L. Cohen(1978),"TheUseofSub- http://www.ft.com].
jectiveExpectedUtility Theoryas an Aid to Understanding Wiener, Joshua L. andTabitha A. Doescher (1991),"AFramework
Variables ThatInfluence HelpingBehavior," Journal ofPer- forPromoting Cooperation," Journal ofMarketing, 55 (April),
sonality andSocialPsychology, 36 (10), 1138-51. 38-47.
Mahoney, John(1976),"TheRelationship ofAnticipated Effec- and (1994),"Cooperation and Expectations of
tiveness, Alienation, andValueStructure toPlanned Participa- Cooperation," Journal ofPublicPolicy& Marketing, 13(Fall),
tionina National MeatBoycott," Psychology, 13(2),39-47. 259-70.
McMillen, DavidL., Deborah Y. Sanders, andGaryS. Solomon Witkowski, Terrence F. (1989),"ColonialConsumers in Revolt:
(1977),"Self-Esteem, Attentiveness, and HelpingBehavior," BuyerValuesandBehavior During theNonimportation Move-
Personality andSocialPsychology Bulletin,3 (2),257-61. ment,1764-1776," Journal of Consumer Research,16 (2),
Midlarsky, E. (1984),"Competence andHelping: NotesToward a 216-26.
Model," inDevelopment andMaintenance ofProsocial Behav- Wolman, Leo (1916),TheBoycott inAmerican TradeUnions. Bal-
ior,E. Staub, D. Bar-Tal, J.Karylowski, andJ.Reykowski, eds. timore:Johns Hopkins University Press.
NewYork:Plenum Publishers,291-308.

Consumer
Boycott 1109
Participation

This content downloaded from 131.211.208.19 on Fri, 27 Sep 2013 08:34:15 AM


All use subject to JSTOR Terms and Conditions

You might also like