Ethical Consumerism in Global Perspective: A Multilevel Analysis

This article discusses the determinants of ethical consumerism, which is the deliberate purchase or avoidance of products for political, ethical, or environmental reasons. The study examines interactions between country-level affluence and individual-level predictors of ethical consumerism. The results highlight the importance of interactive models of ethical consumerism and of political participation more generally.

Ethical Consumerism in Global Perspe
A Multilevel Analysis of the Interacti
between Individual-Level Predictors
Country-Level Affluence
Nik Summers
Indiana University
ABSTRACT
Early empirical research on ethical consumerism—the deliberate purchase, or avoida
of products for political,ethical,or environmental reasons—was primarily individualistic in
nature.Recently,scholars have demonstrated the importance of structural and cultural con-
texts to the explanation of ethical consumerism,rendering explanations that fail to account
for such contextsincomplete.Unfortunately,mostof thisresearch hasbeen contained
within Europe,limiting potentially importantcountry-levelvariation.Because theories of
ethicalconsumerism suggest interactive relationships between individual- and macro-lev
variables,the Euro-centric nature of existing research raises questions about theoretical g
eralizability across all levels of analysis.This study uses the 2004 citizenship module of the
InternationalSocialSurvey Program (ISSP)—a data set that allows for increased country-
level heterogeneity while maintaining the highest standards of data quality—to run a se
of multilevel,logistic regression models with cross-levelinteractions between country-level
affluence and individual-levelpredictors.Seven of the eight individual-levelpredictors ana-
lyzed in these interactions are either more influentialin high-affluence countries than in
low-affluence countries or exhibit statistically uniform effects across the range of afflue
The lone exception is association involvement,which is more influentialas affluence de-
creases.The need to develop interactive models of political participation is discussed.
KEYWORDS : ethicalconsumerism;interaction effects;association involvement;low-cost
hypothesis; cross national.
Although sustained scholarly focus on ethicalconsumerism is recent,people across the world have
long called upon their political,ethical,and moral beliefs when making decisions in the marketp
Colonists,in what would later become the United States,forged a collective politicalidentity out of
the shared experience of boycotting British goods (Breen 2004).Severaldecades later,abolitionists,
The author thanks Patricia McManus,Fabio Rojas,Joseph DiGrazia,Michael Vasseur,and the anonymous Social Problems reviewers
for comments on this article. The author also wishes to thank Mannheim University for hosting him during the co
ticle as well as the participants of the Politics,Economy,and Culture workshop at Indiana University for the opportunity to rec
feedback on early versions ofthe project.Direct correspondence to:Nik Summers,1020 E.Kirkwood Ave.,Ballantine Hall744,
Indiana University, Bloomington, IN 47405-7103. E-mail: [e-mail].
VC The Author 2016. Published by Oxford University Press on behalf of the Society for the Study of Social Problems.All rights reserved.
For permissions, please e-mail: [e-mail]
303
Social Problems, 2016,63, 303–328
doi: 10.1093/socpro/spw009
Article
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

including free blacks living in the North,opened “free produce” stores as an alternative to slave-
products and encouraged boycotting of the latter and the purchase of the former (Glick
In the 1930s, the Nazi party encouraged people to buy from German-only stores and fo
to hang “German Shop” signs in their front windows (Friedman 1996).In South Korea,the origin of
consumer goods,and especially the difference between domestically produced goods and t
in Japan,carry a symbolic importance that people carry with them to the register (Nelson
And, in Turkey, everyday citizens make choices to not purchase Coca-Cola products in
what they see as American cultural and economic imperialism (Sandıkcı and Ekici 2009
sumerism,as these examples show,covers a range ofbehavior including the deliberate purchase of
consumer products to reward companies for positive behavior (buycotting) and the del
tion of products to punish companies for objectionable behavior (boycotting).
In contemporary times, ethical consumerism has become a routine part of shopping
in many countries,and markets for organic,fair trade,and eco-labeled products have been growing
for overa decade (EcolabelIndex 2014;FairTrade Foundation 2014;FairTrade USA 2012;
Organic Trade Association 2011,2012; Steering Committee of the State-of-Knowledge Assessm
of Standards and Certification 2012).Grocery stores are now littered with products that promise
vironmental protection and, to a lesser extent,the fair treatment of poor farmers and factory worke
Certification ofethicalproduction by non-governmentalorganizations and accompanying labeling
initiatives has blossomed to become a major global effort across many industries.This “transnational
private regulation” (Bartley 2007) is intended to influence production practices across g
chains(Gereffiand Korzeniewicz1994)—somethingthathasproven difficultfor territorially
bounded nation-states (Djelic and Sahlin-Andersson 2006).Although allof this activity has resulted
in less demonstrable change in production practices than advocates have hoped for (B
2015),it is clear that ethical consumerism has become a common way that people are a
influence the “politics of products” (Micheletti2003).Some scholars argue it is analogous to social
movementactivity when the decisions ofmany individualconsumers aggregate up to a kind of
“individualised collective action” (Micheletti 2003).
A considerable amount of research has been devoted to studying the determinants o
erism in the last decade,most of which has been focused at the individual level,and much of which has
used data from a single country (Andorfer 2013; Micheletti and Stolle 2005; Strømsnes
2005).Drawing on comparative work that emphasizes the sociostructural contexts of pol
participation (Huckfeldt 1979; Schofer and Fourcade-Gourinchas 2001),a few important studies have
investigated the importance of context in providing opportunities for and constraints on
erism (Koos 2011,2012;Neilson and Paxton 2010;Stolle and Micheletti2013;Thøgersen 2010).
People do not choose whether to ethically consume in a vacuum; instead, their behavio
characteristics of the regions and countries they live in (Jacobsen and Dulsrud 2007).
These accountssuggestan interactive modelof ethicalconsumerism in which the effectof
individual-levelpredictorsvariesacrosscountry ascountry-levelcharacteristicsvary.To testthis
model,some studies have run cross-level interactions between individual- and country-le
but these analyses have been limited to the European context.As such,the influence ofcontext is
probably underestimated and important interactions between individual- and country-levariables
may be obscured by limited country-level heterogeneity (Thøgersen 2010).This study addresses this
imbalance in existing research by including overa dozen countriesfrom the GlobalSouth and
Eastern Europe,as called forrecently by leading scholarsof ethicalconsumerism (Stolle and
Micheletti2013:271) Because country-levelaffluence is the primary contextualfactor in structuring
opportunities for and constraints on both boycotting and buycotting (Koos 2012),this study exam-
ines interactions between it and the established individual-levelpredictors ofethicalconsumerism.
The most useful theoretical tool for understanding these interactions has been the low-
sis,which states that individual-levelvalues and attitudes are influentialon behavior only when con-
texts render the costs of behavior low.
304 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

The inclusion of non-Western countries alerts scholars to the dangers of theoretical
zation and has practicalimplications for the development of ethicalconsumerism beyond the West.
These are important issues as ethical consumerism is both an important type of politic
self and an important pathway to broader political consciousness,especially for marginalized popula-
tions (Stolle and Micheletti2013).The results of this study highlight the importance of interac
modelsof ethicalconsumerism and ofpoliticalparticipation more generally.Generalizing from
single-country models,or even from models that use multicountry aggregated data,can lead to in-
complete or even plainly wrong conclusions about the determinants of participation.
E T H I C A L C O N S U M E R I S M A S P O L I T I C S
Although many scholars have accepted ethical consumerism as a meaningful form of p
ment, others remain skeptical.Critics charge companies with “green-” and “fair-washing” and
icaland environmentally responsible production have become commoditized.Gay Seidman (2007)
has documented a race-to-the-bottom in productcertification and labeling,as companies work to
find the least costly certification that enables them to claim ethical production.In addition to the wa-
tering down of standards, observers see in this process a cynical ploy to appeal to the
of wealthy consumers willing to spend more for conspicuous specialty products,not a plausible way
to significantly alter production practices of polluting or exploitative companies (Goodm
Moreover,ethicalconsumerism often occurs when individuals,alone in grocery store aisles,at-
temptto weigh information presented on confusing ormisleading productlabels (Bostro¨m and
Klintman 2008).It is hard to see how consumers,acting on the basis of such information,can act as
sophisticated regulators of global capitalism; instead, there is evidence that consumer
bels heuristically,as rough guarantees of responsible production (Adams and RaisboroughIn
this attempt to turn consumers into regulators through product information (Schneibe
2008),critics see a dangerous “individualization of responsibility” for collective problem
tinuation ofthe neoliberalturn from governmentto governance (Bevirand Trentmann 2007;
Maniates 2001; Maniates and Meyer 2010).The concern is that consumers,now burdened with the
responsibility for global problems when making purchases (Barnett et al.2011; Sassatelli 2007; Stolle
and Micheletti2013),view their job as done once the purchase is made (Smith 1998).Ethicalcon-
sumerism may lead to the atrophy of collective action if it induces a false sense of poliefficacy
(Szasz 2007).
Furthermore,the goals of corporations are profit and growth not the protection of the e
ment or the rights of workers,and most industries are characterized by intense competition a
pressure (Dauvergne and Lister 2013).In such an environment,good intentions can be swept aside
by competitive pressures from companies thatlack commitmentto ethicalprinciples.As a result,
there isskepticism thatthe capitalistmarketis institutionally suited to delivering socialchange.
Instead,government regulation is seen as the only force strong enough to fundamentall
duction practices; effort expended elsewhere is “frolic and detour” (Reich 2008:14).Ethical consum-
erism as a means of politics appears even more dubious when one considers the size o
for labeled products and the willingness of consumers to pay associated price premiumEven sup-
porters are forced to acknowledge that markets for labeled products are niche and the
suggests a small minority of the population is willing to pay more for guarantees of eth
(Bartley et al. 2015:ch. 2; Hainmueller and Hiscox 2015; Hainmueller, Hiscox, and Seq
One response to these critiques is to acknowledge that ethical consumerism can be
and parochial,butto pointoutthatmuch conventionalparticipation is as well(Schudson 2007;
Willis and Schor 2012).The idea that democratic politics is guided by civic-mindedness whi
sumer behavior is oriented to narrow self-interest does not hold up to scrutiny.Instead,market be-
havior is “utterly saturated” (Malpass et al.2007) with morality (Zelizer 2009) and is embedded
social roles and relationships (Barnett et al.2011; Clarke 2008; Granovetter 1985),while democratic
Ethical Consumerism in Global Perspective305
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

politics can be viciously competitive and self-interested.Is voting to keep one’s personaltaxes low
any more civic-minded than purchasing fair trade coffee?
As for concerns that ethical consumerism diverts attention from other types of particthere
is little evidence that it does so.Instead,across small- and large-sample studies the evidence po
clearly in the direction of compatibility between ethical consumerism and other types o
(Clarke et al.2007;Connolly and Prothero 2008;Neilson and Paxton 2010;Stolle,Hooghe,and
Micheletti2005;Willis and Schor 2012) and supports a view ofethicalconsumerism as “one in a
toolkit of [political] actions” (Willis and Schor 2012:166) and an expansion of the partic
ertoire (Micheletti 2003).
Another response is to point out that there are important indirect effects of ethicalconsumerism
that are often overlooked by critics.Clive Barnett and colleagues (2011) show that ethical consu
ism provides organizations with rhetoricalresources to make claims on nation-states;in struggles
with legislatorsand otherstakeholders,the demand forlabeled productsbecomesevidence for
broad-based desire for legislative action.In this way,and in others,ethical consumerism allows multi-
national corporations,transnational governance organizations,and nation-states to engage in politics
in ways that would not be possible if limited to conventional channels (Micheletti 2003)
directeffectis the potentialthatethicalconsumerism has to awaken politicalconsciousness that
would otherwise lie dormant.When consumption is politicized people have fodder for the dev
ment of new political identities (Shaw,Newholm,and Dickinson 2006) as the purchase of products
becomes “proof of the importance of their [political] aspirations” (Sassatelli 2006:221).Ethicalcon-
sumerism acts as a venue for the exploration of new politicalidentities that are subsequently “trans-
lated into rights and then become the basis of political action” (Hilton 2003:315).
These “venues for action” (Micheletti2003) may be especially important for women,racialand
ethnic minorities,and other groups marginalized within conventionalpolitics.Ethicalconsumerism
has amplified the political voice of African Americans during the American Revolution (
the abolitionist movement (Glickman 2004),the Great Depression (Greenberg 1999),and the civil
rights movement (Cohen 2004).Excluded from conventionalpolitics,even unable to cast votes for
representation,women,and especially housewives,have been at the center of many buycott and bo
cott campaigns throughout history (Davies 2001; Hilton 2002; Sassatelli2007).Boycotting has also
been famously recognized as a “weapon ofthe weak” whether itis used by poor Malaysian field-
workers to resist the power of plantation owners (Scott [1985] 2008) or by poor Irish re
with the behavior of English landlords (Friedman 1999).
Beyond the West
Because consumer power has historically been centralized in the developed economies
industrialized world mostof the demand-side focus on ethicalconsumerism has been on Western
Europe and the Anglo offshoots.People with fewer “financialdegrees offreedom” and who live in
countries with economies that provide a low assortment of goods have less opportunity
sume (Koos 2012:40). Why, then, study ethical consumerism beyond the West?
For one, the emergence of significant consumer demand around the world means tha
labeled products,even when such products are more expensive,are no longer limited to the West
(Hobson 2004;Klein 2009;Stolle and Micheletti2013:ch.8). Non-Western countries,across the
spectrum ofaffluence,are now a significantpartof transnationallabeling initiatives such as the
Global Ecolabelling Network (GEN 2014) and rates of ethicalconsumption meet or exceed 10 per-
cent in numerous non-Western countries in the InternationalSocialSurvey Program (ISSP 2004)
sample,as displayed in Figure 1.Moreover,some forms ofethicalconsumerism are low cost.One
can engage in ethicalconsumerism simply by consciously rejecting one of two similarly price
natives in favor of the other,as occurs in cases of politically motivated brand rejection (Sandık
Ekici 2009).
306 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Also,as discussed previously,ethical consumerism has the ability to generate and nurture p
consciousness.So,even ifabsolute rates ofethicalconsumerism are low in non-Western countries
the politicization of market behavior could have broader consequences.In China,private home own-
ership,for many the most significant act of consumption,fundamentally altered the relationship be
tween the Chinese state and its citizens and gave consumers a new sense of politicalefficacy (Davis
2000,2006).If politicalidentities develop through ethicalconsumerism,and if ethicalconsumerism
is a form of politicalengagement suited to the incorporation of the marginalized,then it may be an
important site of entry to politics for poor peasants,factory workers,women,and people facing non-
democratic governments around the world (Scott [1985] 2008; Yan 2000) This does no
calconsumerism would necessarily make globalpolitics more democratic,as it can be nationalistic
and exclusionary (Frank 2000).Nor does it mean that ethical consumerism would topple auth
ian regimes or create a globalcitizenship capable ofregulating capitalistproduction,as advocates
0 10 20 30 40 50 60
0 5000 10000 15000 20000 25000 30000 35000 40000 45000
Philippines
Venezuela
Brazil
Uruguay
Bulgaria
Russia
Latvia
Chile
Mexico
Poland
Slovakia
Hungary
Czech Republic
Portugal
South Korea
Slovenia
Israel
Cyprus
New Zealand
Taiwan
Spain
Finland
Japan
Germany
France
Great Britain
Belgium
Sweden
Denmark
Australia
Austria
Netherlands
Canada
Switzerland
Ireland
United States
Norway
Percent Ethical Consumers
GDP/Capita
GDP/Capita
% Ethical Consumers
r = .73
(in 2013 international dollars)
Figure 1. Affluence and Ethical Consumerism by Country
Ethical Consumerism in Global Perspective307
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

hope.But if ethicalconsumerism is another pathway to political consciousness and participaes-
pecially for those marginalized within global political processes, then it is an expansion
repertoire for those that most need it.
P R E D I C T O R S O F E T H I C A L C O N S U M E R I S M
The empirical focus of this article is testing the interactive nature of the individual- and
determinants of ethical consumerism. At the individual level, ethical consumerism is m
mographic characteristics,resources such as income and education,social capital,and “motivational”
factors such as values,norms,and attitudes.At the contextuallevel,country-levelaffluence is espe-
cially powerfulbut other macro factors are important as well,including additionaleconomic charac-
teristics,politicalstructures,culture,and socialcapital.Together,these contextualforces represent
the opportunity structures (McAdam,McCarthy,and Zald 1996) for ethical consumerism that pow
erfully shape individual-level behavior.
Individual-Level Predictors
Demographic Characteristics
Most studies find that women ethically consume at higher rates than men (Ferrer and F
Forno and Ceccarini 2006; Koos 2012; Micheletti and Stolle 2005; Neilson and Paxton 2
and Micheletti 2006, 2013; Strømsnes 2005; Tobiasen 2005; Yates 2011). Although this
attributed to the idea that women shop more,there is evidence that the gender effect persists ev
when amount ofshopping is controlled for (Michelettiand Stolle 2005).More likely explanations
are that women tend to be involved at higher rates in voluntary associations,such as animalrights
groups,that prime people for ethicalconsumerism or that women are more likely to hold attitud
and values consistent with the protection of the environment or labor rights (Stolle and
2006).Ethical consumerism may also be more attractive to women than more conventio
politics for severalreasons,including that it lacks a hierarchicalparticipatory structure and member-
ship requirements(Marien,Hooghe,and Quintelier2010;Stolle and Hooghe 2011;Stolle and
Micheletti 2006).Age has also been included in most previous studies,though typically as a control
variable without much in the way of interpretation.
Resources for Participation
Education is the most consistent individual-levelpredictor of ethicalconsumerism (see studies cited
above for gender finding).This is not surprising given the overwhelming and positive associat
education with other types of political participation (Kam and Palmer 2008; Rosenstone
1993; Verba, Schlozman, and Brady 1995) and is typically explained through a simple r
(Brady,Verba,and Schlozman 1995) that argues education provides people with cognitiveinforma-
tional,and motivational resources to act.Understanding the politics of products and sorting throu
the claims made on product labels can be daunting tasks,and education provides people with the
skills to complete them.Education may also provide people with the requisite culturalcapitalto
know how to use their consumption decisions to signalsuperior socialstatus and,thus,it may con-
tribute to class-based,symbolic exclusion (Andorfer 2013;Johnston 2008;Johnston and Baumann
2010).
Results for household income have been uneven.Sebastian Koos (2012) argues this inconsistenc
is a result of the conflation of boycotting and buycotting and demonstrates that income
predictor ofbuycotting but not boycotting when they are disaggregated,using data from the 2003
European SocialSurvey (ESS).Lauren Copeland (2014),however,using data from the United
States,obtains different results when disaggregating the behavior even further.She finds that income
is not usefulin distinguishing boycotters or buycotters from those that do not ethically co
while itis usefulin distinguishing “dualcotters” from “nocotters.” To furthercomplicate matters,
308 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Lisa A.Neilson (2010),using ESS data,finds that income is useful in distinguishing boycotters,buy-
cotters,and dualcottersfrom nocotters.More research isneeded to untangle thiscomplicated
relationship.
Social Capital
Membership in voluntary associations,whether explicitly political or not,is predictive of political par-
ticipation ofvarious types (Almond and Verba [1963] 2015;Leighley 1995;Verba and Nie 1972)
and thisrelationship holdsfor ethicalconsumerism (Neilson 2010;Neilson and Paxton 2010).
Associations increase interest in political matters (Andrews et al.2010),generate feelings of trust and
reciprocity (Fukuyama 1995; Putnam 1994),and provide opportunities for the development of ci
skills that facilitate engagement (Baggetta 2009). With respect to ethical consumerism
volvement supplies people with information to make consumption decisions and motiv
on that information (Neilson and Paxton 2010).Associations increase access to information by in
creasing the number of weak ties that people have—relationships that are particularly
transmission of novel information (Granovetter 1985; Mutz 2002; Neilson 2010; Warde
also increase normative pressure to act (Andrews et al.2010; Clarke et al.2007) by exposing people
to visible consumption norms and sanctions tied to norm violation (Neilson and Paxton
General trust is thought to increase rates of ethical consumerism as it creates an ex
iprocity and alleviates concerns people have thatothers willfree ride on their efforts (Nannestad
2008).Trust in political institutions is expected to be negatively correlated with ethical c
because if people believe in the efficacy of conventional politics they may be less likel
politicalenergies toward the market (Stolle et al.2005).Cross-nationalresearch finds evidence in
support of both of these expectations (Koos 2012; Neilson and Paxton 2010).
MotivationalFactors: Values,Norms,and Attitudes
Values, norms, and attitudes are viewed as “motivational” factors that are crucial to u
decision to ethically consume (Andorfer 2013;Andorfer and Liebe 2013;Koos 2012;Shaw etal.
2005;Sunderer and Ro¨ssel2012).Ronald Inglehart’s (1990,1997) work on postmaterialism has
been especially prominent here, as scholars have found strong evidence that postmat
influential predictors of ethical consumerism (Stolle et al.2005).As societies develop economically a
macro-levelvalue shift from materialism to postmaterialism occurs,quality of life issues such as hu-
man rights gain prominence,and people become more likely to ethically consume as a resultOther
values are important as well,including solidarity and concern for the environment (Andorfer 2
independence and equality (Shaw et al.2005),and self-transcendence (Koos 2012).More proximal
to ethical consumerism itself,holding positive attitudes toward fair trade and feeling a person
gation to help through one’spurchasesare also importantpredictorsof ethicalconsumerism
(Andorfer 2013; Andorfer and Liebe 2013; Sunderer and Ro¨ssel 2012).
Contextual Predictors
Scholars of ethical consumerism have recently applied insights from the opportunity s
ture (McAdam et al.1996) to describe how characteristics of countries and regions constra
able individual-levelethicalconsumption (Koos 2011,2012;Neilson and Paxton 2010;Thøgersen
2010; Wahlstro¨m and Peterson 2006).The idea is quite simple: individual-levelconsumption deci-
sions occur within contexts that powerfully shape resource distributions,opportunities to ethically
consume,and cultures ofconsumption and politicalengagement.Economic,political,and cultural
characteristics have all been identified as important, as well as contextual social capita
Economic characteristics are thoughtto be especially importantfor providing households with
enough resources to ethically consume and enough opportunities to do so.More specifically,impor-
tant economic factors include retailing and price structures,the supply oflabeled goods,aggregate
Ethical Consumerism in Global Perspective309
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

demand for products, and affluence (Koos 2012; Thøgersen 2010). Country-level afflue
proxy for level of economic development and is important in influencing many other as
tion’s economic structure.Moreover,affluence raises the financialposition ofall households in a
country relative to similarly positioned households in less-affluent countries,thereby increasing the
freedom people have with which to make purchasing decisions.Politicalcharacteristics are believed
to influence how citizens engage in politicaland civic life by institutionally encouraging some typ
of engagement and discouraging others.For instance,polities with strong,centralized states discour-
ageassociation membership relativeto less“statist”politicalsystems(Schoferand Fourcade-
Gourinchas2001).Importantpoliticalfactorsfor ethicalconsumerism include the institutional
characteristics of the state,the character of associational life in a country,the nature of regulation rel-
evant to ethicalconsumerism,and the levelof state-involvement in labeling schemes (Koos 2012
Thøgersen 2010). Globalization has also been included in previous studies in political, eand
cultural varieties and is thought to increase ethical consumerism since it is often theori
ing conventional political institutions (Koos 2012).
Cultures of consumption and production vary across region and country and these cu
patterns ofethicalconsumption.For instance,the purchase oflabeled products may be depressed
around the Mediterranean in part because of the historicalfocus on the protection of localand au-
thentic cuisines instead of organic or fair trade consumption (Grasseni2003; Sassatelliand Davolio
2010; Sassatelli and Scott 2001). Shared norms and values at the contextual level,such as postmateri-
alism or buying green,also influence patterns of ethicalconsumerism at the individuallevelby en-
couraging compliance and imposing sanctions on norm violators.For example,the poor are often
stigmatized for eating “fast food” and shamed for their eating habits (Guthman 2007).Lastly,macro-
level social capital is important in influencing individual patterns of ethical consumerismIndividuals
in high-socialcapitalsocieties are more likely to ethically consume because they have great
to information and more motivation to act as a result of access to more trusting and int
networks. Neilson and Pamela Paxton (2010) find that regional-level generalized trust i
itively associated with ethical consumerism.
I N T E R A C T I O N S B E T W E E N I N D I V I D U A L - L E V E L A N D
C O U N T R Y - L E V E L V A R I A B L E S
Because socialcontext powerfully shapes patterns of ethicalconsumerism even among the relatively
homogenous countries of Europe,the introduction of non-European countries could mean subs
tial differences in the explanatory significance of individual-level predictors.Studies so far have unani-
mously confirmed the positive association between education and ethical consumerismbut there is a
wide variety of educationalsystems around the world (Clark 1986).Does the finding for education
hold up when non-Western and poor countries are included? The same could be asked
lightof variation in gendernorms(Hunter,Hatch,and Johnson 2004;Iversen and Rosenbluth
2006), or for any of the other individual-level variables.
Unfortunately,there are currently more questions than answers or even articulated hyp
Existing accounts of ethicalconsumerism imply the presence of interaction effects,but infrequently
elaborate expectations with any specificity.Veronika A.Andorfer (2013),in advocating theoretical
synthesis through a macro-micro-macro model (Coleman 1990),highlights the need for “bridge the-
ories” to specify how macro contexts influence micro behavior.The most developed expectations for
interaction that exist in the literature come from the “low-cost hypothesis,” an influent
of rational choice theory that focuses rather specifically on values and attitudes.
The foundation of the low-cost hypothesis is a rational choice model of human behav
with the premise that people are utility maximizers and that they make decisions based
costs of actions (Liebe and Preisendo¨rfer 2010).In short,people choose behaviors that yield them
the greatestutility atthe lowestcost.This “narrow” understanding ofhuman behavior has been
310 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

augmented by “wider” models,which add cultural and social variables to the economic formal
narrow accounts.With respect to low-cost approaches,these additions have come mostly in a focus
on how supra-individual contexts raise and lower the costs of behaviors and,thus,constitute a “struc-
ture of opportunity” for individual action (Ro¨ssel 2008) and “boundary conditions” for
of attitudes and values on behavior (Guagnano, Stern, and Dietz 1995).
According to this model,as externalcontexts raise the costs of a behavior,individual-levelvalues
and attitudes become less important in explaining that behavior; only in low-cost situa
matter.In very high-cost situations,in fact,economic considerations become “dominant decision
teria” (Diekmann and Preisendo¨rfer 2003:446) and “psychological characteristics” of
come almost irrelevant . . . Economics becomes essentially autonomous” (Schons, Resand Wieseke
2010:126).In low-cost situations that represent little threat to an individual’s material we“it
is easierfor actorsto transform theirattitudesinto corresponding behavior”(Diekmann and
Preisendo¨rfer 2003:443).Evidence in support of the low-cost hypothesis has been found in li
research (Ro¨ssel2008),studies ofenvironmentalconcern (Diekmann and Preisendo¨rfer 2003),in-
cluding recycling behavior (Derksen and Gartrell 1993; Guagnano,Stern,and Dietz 1995),criminol-
ogy (Kroneberg,Heintze,and Mehlkop 2010),and in experimentalgame scenarios involving real
money in high stakes situations (Schons etal.2010).With respectto ethicalconsumerism,Koos
(2012) finds that self-transcendence values,which are closely related to postmaterialist values,are
more influential in low-cost contexts, as expected.
Low-cost theories lead to the following hypothesis: The influence of individual-level a
values willbe greater in low-cost (high-affluence) countries.In other words,there willbe a positive
interaction effect between individual-level attitudes and values and country-level afflu
There is less guidance for what to expect in the way of interactions between countryafflu-
ence and other individual-levelvariables such as gender or education.The low-cost hypothesis does
not clearly specify how individual-leveldemographic characteristics or resources might interact
contextual costs,beyond the general sense that very high-cost contexts reduce all non-eco
siderations to insignificance.Linda Derksen and John Gartrell’s (1993) study of recycling beha
Alberta,Canada,providesa partialexception to this.Working from a broader,and lesssocial-
psychological,understanding of the low-cost approach,they argue that social contexts raise or lowe
the effort that individuals need to expend to engage in specific behaviors and that ind
iables other than attitudes and values might also become more influentialin low-cost situations as a
result.As such,they expected to find interaction effects between age,education,income,and job
prestige with belonging to a blue-box household (their low-cost condition),even though multicolli-
nearity prevented them from testing all of their expectations.
Moving beyond the low-costhypothesis,studies have examined cross-levelinteractions between
contextualopportunity structures and non-electoralpoliticalparticipation (Braun and Hutter 2014;
Marien et al. 2010; Vrablıkova 2013) and attempts at general, multilevel, theoretical in
cently been made (Barrett and Brunton-Smith 2014).However,ethical consumerism has not been in-
cluded in these studies in a significant way.Indeed,Katerina Vrablıkova (2013:10) explicitly mentio
the lack of theoreticalexpectations for ethicalconsumerism as a reason for not including ethicalcon-
sumerism measures in her cross-levelanalyses.In addition to providing an expanded test of the low
cost hypothesis,then,this study willprovide a broad exploratory analysis of the interactions be
individual-level predictors of ethical consumerism and country-level affluence and, thethe first test
of whether the individual-levelmodelof ethicalconsumerism developed in the West is generaliza
outside of that context. The theoretical implications of these findings are discussed in
DATA AND METHODOLOGY
The InternationalSocialSurvey Program’s (ISSP) 2004 citizenship module supplies the indiv
leveldata for this study.The ISSP is one of the largest (totalN ¼ 52,550 across 38 countries) and
Ethical Consumerism in Global Perspective311
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

most respected sources of data for performing cross-nationalanalyses,especially when non-Western
nations are of interest.As of 2004,more than a dozen non-Western countries supplied data to
ISSP,including countries in Eastern Europe,Asia,and Latin America.This allows for the analysis of
individual-level data from a much wider array of countries than previously studied by sc
cal consumerism.All models presented in the text were run using data from 36 countries w
estimation sample of N ¼ 37,408,with 29 percent missing data.See Figure 1 for a list of the coun-
tries included in the final estimation sample.Taiwan is presented in Figure 1 for illustrative purpos
but is not included in final analyses because of missing data on country-level measuresKey analyses
were also run after using several strategies to reduce missing data, including running m
forming multiple imputation of missing values using chained equations; results were ro
strategies.Also included in my analyses are country-level measures obtained from severaIn
addition to specifications discussed in the text,alternative specifications were tested for most var
ables. Results were robust to specification unless otherwise noted.
Dependent Variable
The dependent variable used in this study is a self-reported measure ofethicalconsumerism.The
measure comes from a series of questions about political participation that begin with t
Here are some different forms of political and social action that people can take.Please indicate,
for each one,whether you: (1) Have done it in the past year; (2) Have done it in the mor
tant past; (3) Have not done it but might do it; and (4) Have not done it and would ne
The wording of the ethical consumerism item is “Boycotted,or deliberately bought,certain products
for political,ethicalor environmentalreasons” (ISSP 2004).Categories (1) and (2) were collapsed
to create a dichotomized measure with people that report any history of ethical consum
as 1 and those that have not ethically consumed coded as 0.
This measure contains both boycotting and buycotting and,thus,is conceptualized as a general
“ethicalization” of marketplace decisions.Although there is some evidence that boycotting and bu
ting are distinct enough from one another to be motivated by a different set of factors
Neilson 2010;Yates 2011) severalstudies analyze dependent variables that combine the two (
and Fraile 2006; Neilson and Paxton 2010; Stolle et al.2005; Stolle and Micheletti 2013) and there ar
methodological and theoretical reasons to question strong arguments for their separati
Theoretically,Monroe Friedman (1996,1999) has argued that boycotting and buycotting are
ten two sides of the same coin,something that even those that argue for treating them separa
knowledge (Neilson 2010:218).For instance,Friedman (1999) discusses efforts by labor unions t
encourage people to purchase products with a union label.Implicit in those efforts is a rejection,or
boycott,of products without the label.Often the designation of a decision or behavior as boycot
or buycotting comes down to a point of emphasis. As he says:
Thus, for example, a proposed consumer boycott by Americans of Japanese goods m
fined as a consumer buycott of American goods,the change of emphasis from the negative to
the positive reassures Americans that they are being asked to act patriotically rather
vinistically (p. 445).
Similarly,the decision of where to purchase lunch might involve a simultaneous rejection
tized fast food in favor of the deliberate purchase of “good” food supplied by the local f
(Guthman 2008; Johnston,Szabo,and Rodney 2011).It is not clear whether such behavior should
be classified as boycotting,buycotting,or both.Furthermore,and perhaps more to the point,it is not
clear how people responding to survey questions think such behavior should be categoIt is
clear, though, that such behavior is an “ethicalization” of market behavior.
312 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Most large-scale surveys also employ opaque measures that make determining the
tween boycotting and buycotting difficult.For instance,Copeland (2014) sets out to investigate the
different social motivations of boycotting and buycotting by using two separate measu
cotting and one of buycotting.Unfortunately,her buycotting measure conflates the two behaviorIt
reads, “In the past 12 months, did you purchase one product or brand over another fo
icalreasons?” (Copeland 2014:179,emphasis added).It is easy to see the dualmotivations that
Friedman notes in many purchasing decisions reflected in that question.Similarly,Neilson (2010)
finds the questions she uses from the European SocialSurvey problematic insofar as her buycotting
question specifies that the respondent purchased a product deliberately for “political,ethical,or envi-
ronmentalreasons” whereas the boycott measure does not mention such motivations aninstead,
simply reads“boycotted certain products.” Survey measures available forcross-nationalresearch,
then,more readily lend themselves to a combined measure such as the one used in this
studying the potentially different social foundations of the behaviors.
Individual-Level Measures
Several demographic and resource measures are included: gender,education,and household income.
Gender is included as a binary variable with women equalto 1 and men as the reference category.
Education is also a binary variable with those that have completed a college degree co
others as the reference.Household income is coded following Koos’s (2012) decision to use a
dardized measure thatplacesan individual’searningsin relation to hercountry’smean income
(CMI). The measure used here is a five-category measure composed of people whose h
come is less than 50 percent of CMI,between 50 and 80 percent,between 80 and 120 percent,over
120 percent, and those with missing data for income.
Socialcapitalis operationalized with three variables:association membership,generaltrust,and
trust in institutions.Association membership is included as a summed index of membershi
types of associations: (1) politicalparties,(2) trade unions,(3) church or other religious organiza-
tions,(4) sports,leisure,or cultural groups,and (5) a catch-all category for any other voluntary a
ciation.Each originalmeasure offered respondents four response categories:(1) “Belong actively,
participate,” (2) “Belong,don’t participate,” (3) “Used to belong,” and (4) “Never belonged.”
chotomized measure was created for each ofthe five types ofassociations where categories two to
four are equal to 0 and category one remains coded as 1. The index is the sum of thos
with values ranging from 0 to 5.General trust and trust in political institutions are included as
continuous,likert-style measures with higher responses aligned with a theoreticalexpectation of in-
creased likelihood of ethical consumerism.
A measure of the attitude that it is important to participate in politics is included in a
replicate the finding that personalnorms influence ethicalconsumerism (Andorfer 2013;Andorfer
and Liebe 2013). The measure is the sum of two individual,seven-item measures: the first,a measure
of how important for good citizenship the respondent believes it is to always vote in eland
the second how important they consider active association membership for good citizeA proxy
measure of self-transcendence values is also included that is the sum of three items: t
tolerance toward others, and the second and third,respectively, capture the respondent’s belief th
is important to look after the welfare of proximal and distant others.
Two additionalindividual-levelcontrols,personalpoliticalefficacy and age,are also included in
analyses. Efficacy is coded as the sum of two items: the first, a measure of responden
the notion ofrespondent influence on what government does and the second a measureagree-
ment with the statement,“I don’t think the government cares much what people like me thin
along with a squared term,is included in models per previous research.Aside from age and the
squared term for age, none of the individual-level variables are correlated above r ¼ .
litical attitude and self-transcendence values).
Ethical Consumerism in Global Perspective313
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Country-Level Measures
All country-levelmeasures use 2003 data when possible to ensure measurement prior to or
with data collection of the ISSP.Aspects of the political,cultural,and economic opportunity structures are
included as well as measures of macro-level social capital and globalization.A logged measure of GDP per
capita obtained from the International Monetary Fund’s (IMF) Economic Outlook databa
primary operationalization of country-levelaffluence.Also included is a country-levelvariable that assigns
each respondent the percentage of postmaterialists in her country.A binary measure of democratization is
included that is a mix ofa country’s democracy score,as measured by the Polity IV project (Marshall
Jaggers,and Gurr 2011) and its history with the current regime.Prior research has shown that citizens
need time to become acquainted with politicalinstitutions for their behavior to become fully adapte
them (Franklin 2004). Countries were coded as 1 if they received the full democracy sc
30 or more years of experience with continuous democratic rule.Other countries were coded as 0.Koos
(2012) uses levelof statism (Jepperson 2002) to capture politicalcontext in Europe,but as the ISSP in-
cludes countries with a substantially wider range of variation in political institutions, a m
is appropriate.A commonly used index of globalization is also included (Dreher 2006).Lastly,the average
association involvement score per country, computed from the individual-level index di
cluded as a country-level measure to control for macro-social capital, following Neilson
A N A L Y T I C S T R A T E G Y
Analyses consist of two main parts: (1) the presentation and discussion of a series of m
regression models that allow for random intercepts by country,and (2) a series of multilevellogistic
regression models that also allow for random slopes to be estimated for individual-leveThe
random intercept models are used to test for replication ofthe findings ofprevious research and to
establish the importance of each individual- and country-level variable included in the fi
random slope models allow for the cross-levelinteractions that are necessary to test the interacti
hypotheses.In other words,it is the random slope models that answer the question ofwhether the
individual-level variables vary in explanatory significance across the range of country-le
The models for my analyses can be written as follows.
Multilevel Logistic Regression Models—Random Intercept
Pr yij ¼ 1 ¼ b1j þ b2X2ij þ b3X3ij þ ::: þ bpXpij þ eij
Where Pryij ¼ 1 is the probability that respondent iin country jis an ethicalconsumer,b1j is a
country-specific intercept term,Xp is an individual- or country-levelcovariate,and eij is an idiosyn-
cratic error term. The intercept term can further be broken down into:
b1j ¼ b1 þ f j
Where b1 is the population mean value for the intercept and fj is the country-specific random com-
ponent that shifts the intercept up or down for each country.
Multilevel Logistic Regression Models—Random Slope
The random slope models are expressed similarly to the random intercept models,but with an addi-
tionalcomponent to represent the estimation of a random slope for individual-levelvariables that is
allowed to vary by country. The model can be written as follows:
Pr yij ¼ 1 ¼ b1j þ b2X2ij þ b3X3ij þ ::: þ bpXpij þ eij
314 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Where the intercept term is still:
b1j ¼ b1 þ f 1j
But one individual-level covariate per model is allowed to vary randomly. This is expre
bpj ¼ bp þ f pj
Where bpj is any individual-level covariate,bp is the population average effect of variable p; and fpj is
the country-specific random component for the effect of variable p.
R E S U L T S
Descriptive Statistics
Figure 1 shows that rates of ethical consumerism pretty closely track country-level afflas prior
research suggests.Most likely,this reflects differences in retailing structures for organic,fair trade,
and eco-labeled products and household income across countries (Koos 2012).To be clear,later
models do controlfor household income,but income is included as a measure relative to count
means.This means thatthe models are controlling for individual-levelincome differences within
countries,not acrosscountries.As Koos(2012) argues,country-levelaffluence shiftsthe entire
income distribution up or down, such that similarly positioned people in poorer countri
financial degrees of freedom than their wealthy-country peers.In an absolute sense,income differen-
cesacrosscountriessurely contribute to differentratesof ethicalconsumerism,but including
individual-levelincome without standardizing it by country would be akin to measures ofcountry-
levelaffluence like GDP percapita,which is already included in the models.In any case,these
descriptive results provide strong prima facie support for contextual theories of ethica
With a few exceptions,the countries with the highestratesof ethicalconsumerism are wealthy
Western nations and the countries with the lowest rates are Eastern and Southern EurLatin
American, and Southeast Asian.
Table1 showsthe descriptivestatisticsfor the aggregatedata.The proportion ofethical
consumers—28 percent—is similar to what other studies have found,though slightly lower than the
34 percent Neilson and Paxton (2010) found using ESS data.Slightly more women than men are in
the sample, less than 20 percent of the sample has finished a college degree, and the
ation involvement mean score is less than one (out of a potential five).The sample is slightly biased
toward those with higher income.
Random Intercept Models
Model 1 in Table 2,an empty multilevel model with no covariates, was run using just the d
variable and a country variable to specify higher-level clustering. This empty model is
pooled logit alternative (LR-Test v2 ¼ 6438.39***,p <.001),providing the first indication of hier-
archically structured data. Furthermore, the residual intraclass correlation (ICC) for the
is .262, which means that about 26 percent of the variation in the intercept is explaine
between countries.This is a larger amount of variance at the country levelthan what Koos (2012)
found using ESS data (17 percent).The greater country-levelheterogeneity included in the ISSP,
then, is reflected in a higher share of unexplained country-level variance.
The resultsfor Model2, the fullindividual-levelmodel,show thatthe effectsfor all of the
individual-level variables are in line with past research,both in statistical significance and direction o
effect.Being a woman,college educated,and having greater than 120 percent of CMI all increase
odds of being an ethical consumer.Age has the curvilinear effect found in past research.Being gener-
ally trusting ofothers,distrustfulof politicalinstitutions,feeling an obligation to participate in
Ethical Consumerism in Global Perspective315
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

politics,scoring high on self-transcendence values,and scoring high in politicalefficacy are allalso
associated with higherlikelihood ofethicalconsumption.Although these similaritieswith past
research are real,they mask important variation in the effect of individual-levelvariables across con-
text that would be missed if no attention were paid to potential interaction effects.
Turning to the country-levelvariables,Model3 shows the results ofthe fullrandom intercept
model,including allindividual-and country-levelvariables.Although each country-levelvariable is
positively and significantly associated with ethical consumerism when included in mode
(not displayed),only GDP per capita and the percent of postmaterialists per country remai
cantin the fullmodel.These findings suggestthatcountry-leveleconomic,political,cultural,and
socialcapitalcharacteristics do significantly contribute to forming opportunity structures f
consumption,but that affluence is the primary factor,especially considering that affluence is though
to drive the development of postmaterialist values.
Also of interest in Table 2 is the reduction in the intercept variation and the ICC acros
Seventy-two percentof the unexplained country-levelvariance in the fullindividual-levelmodel
(Model2),as captured by the ICC,is accounted for by including GDP per capita and the perce
postmaterialist by country (Model4).This is mirrored by a similar reduction in intercept variatio
(78 percent).Model4,including only GDP and percent postmaterialist per country is the pre
model for the random slope models.This is because Model 3,the full model including all individual-
and country-levelvariables,contributes only slightincreases in explained country-levelvariance in
Table 1. Descriptive Statistics for Aggregate Data
Mean/Proportion Std.Dev. Min. Max.
Dependent variable (N ¼ 37,408)
Ethical consumers .28 .45 .00 1.00
Individual-level variables
Female .52 .50 .00 1.00
College .17 .37 .00 1.00
Age 45.35 16.77 15 98
Household income:
< 50% CMIa .22 .41 .00 1.00
50-80% CMI .18 .38 .00 1.00
80-120% CMI .20 .40 .00 1.00
> 120% CMI .25 .43 .00 1.00
Missing on income .15 .36 .00 1.00
Distrust in governmentb 3.23 1.06 1.00 5.00
Association involvement .61 .89 .00 5.00
Generalized trust 2.32 .77 1.00 4.00
Political obligation norm 9.99 2.85 2.00 14.00
Self-transcendence values 16.45 3.64 3.00 21.00
Personal political efficacy 5.31 2.18 2.00 10.00
Country-level variables
GDP per capita (logged) 9.84 .61 7.91 10.66
Postmaterialists by country (%) 16.01 8.66 2.38 34.96
Democratization .44 .50 .00 1.00
Globalization index 75.85 11.21 56.14 92.05
Average association involvement .58 .29 .09 1.30
aCMI ¼ country mean income
bCoded such that higher values indicate greater distrust in government.
316 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Table 2. Multilevel Random Intercept Logistic Regression Models
Model 1
(Empty)
Model 2
(Individual Level)
Model 3
(Full)
Model 4
(Preferred)
Individual-level variables
Female – 1.165*** 1.165*** 1.166***
(.031) (.031) (.031)
College – 2.021*** 2.022*** 2.022***
(.071) (.071) (.071)
Income –
< 50% CMIa – .831*** .831*** .831***
(.037) (.037) (.037)
50% - 80% CMI – 1.007 1.006 1.006
(.043) (.043) (.043)
>120% CMI – 1.290*** 1.288*** 1.289***
(.050) (.050) (.050)
Missing on income – .911† .911† .911†
(.044) (.044) (.044)
Age – 1.037*** 1.036*** 1.036***
(.005) (.005) (.005)
Age2 – .999*** .999*** .999***
(.000) (.000) (.000)
Trust – 1.297*** 1.295*** 1.295***
(.025) (.025) (.025)
Distrust in government – 1.152*** 1.153*** 1.152***
(.016) (.016) (.016)
Association involvement – 1.306*** 1.304*** 1.305***
(.020) (.020) (.020)
Personal engagement attitude – 1.032*** 1.033*** 1.033***
(.006) (.006) (.006)
Self-transcendence values – 1.052*** 1.052*** 1.052***
(.005) (.005) (.005)
Political efficacy – 1.070*** 1.069*** 1.070***
(.007) (.007) (.007)
Country-level variables –
GDP per capita (log) – – 1.893** 2.688***
(.443) (.442)
Postmaterialist (%) – – 1.045*** 1.054***
(.012) (.011)
Level of democratization – – 1.498 –
(.373)
Globalization – – 1.013 –
(.012)
Association involvement – – 1.013 –
(.340)
Intercept (odds ratios) .324*** .010*** .000*** .000***
(.058) (.002) (.000) (.000)
(continued)
Ethical Consumerism in Global Perspective317
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

relation to Model4 and because severalof the country-levelvariables are highly correlated.This
preference is further supported by statisticaltests ofmodelfit (Model3 BIC ¼ 35,577.31,Model
4 BIC ¼ 35,549.99, LR-Test [Model 4 nested in Model 3] v2 ¼ 4.27, NS). With that said, all random
slope models were also run including all country-level measures. Key results did not ch
Cross-Level Interactions
Table 3 displays the key results of this study, the cross-level interactions between the
variables and GDP per capita.Each modeldisplayed in Table 3 uses Model4 in Table 2 as a base,
but to clarify presentation only the estimates for the individual-levelvariable used in the cross-level
interaction are shown.Also,even though the cross-levelinteractions are displayed together,each
comes from its own regression,as there is evidence that running cross-level interactions toget
quickly lead to poorly performing models (Stegmueller 2013).In support of the low-cost hypothesis,
self-transcendence values and the attitude thatpoliticalparticipation is importantshow a positive
interaction with GDP percapita.This means thatindividual-levelattitudes and values are more
important in low-cost (high-affluence) contexts, as the theory predicts.
The results for the socialcapitalvariables are less consistent.Generaltrustdisplays a positive
significant interaction with GDP,but distrust in politicalinstitutions is insignificant.This does not
mean that distrust in institutions is not an important predictor ofethicalconsumerism.Instead,it
means that the effect does not differ in statistical terms across the range of affluence iOf
all the individual-levelvariables,only association involvementdisplays a negative and statistically
significant interaction with GDP.This means that the effect ofassociation involvement is actually
strongerin low-affluence countries.The mixed resultsfor these variablessuggestthe effectsof
socialcapitaldepend on country-levelaffluence,type ofsocialcapital,and the interaction between
the two.
The effects of the resource variables—education and income—are not statistically diff
the range of affluence.Having higher levels of education and higher household income make
consumerism more likely regardless of country-levelaffluence,suggesting the generalizability of the
notion that ethical consumerism is an activity of the privileged. For these analyses, inco
tionalized both as a standard z-score measure and as a binary measure of those that m
120 percent of CMI.
The effect of being female displays a positive interaction with GDP,which shows that gender is a
stronger predictor of ethical consumerism as affluence rises.This suggests that the consistency of the
gender effect in previous research may be an artifact ofthe homogeneity ofthe countries used for
analysis. Country-specific logit models were run that support this interpretation. Being
nificantpredictorof ethicalconsumerism in over60 percentof Western European and Anglo-
offshoot countries but retains significance in only 1 of 21 non-Western countries; includ
in Eastern Europe, Southern Europe, Asia, Latin America, and Israel.
Table 2. Multilevel Random Intercept Logistic Regression Models (continued)
Model 1
(Empty)
Model 2
(Individual Level)
Model 3
(Full)
Model 4
(Preferred)
Intercept variation 1.167 1.122 .220 .247
(.274) (.264) (.053) (.060)
Residual intraclass correlation .262 .254 .063 .070
(.045) (.045) (.014) (.016)
Notes: Standard errors in parentheses; results presented in odds ratios. Final N for all models ¼ 37,408.
Source: 2004 ISSP citizenship module.
aCMI = country mean income
†p <.10 *p <.05 **p <.01 ***p <.001 (two-tailed tests)
318 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Table 3. Random Slope Models with Cross-Level Interactions
Model 5
(Female)
Model 6
(College)
Model 7
(Assoc)
Model 8
(Trust)
Model 9
(Distrust)
Model 10
(Income)
Model 11
(Attitude)
Model 12
(Values)
Main effects: individual level
Female .095*** – – – – – – –
(.059)
College – 1.145 – – – – – –
(.1.214)
Association involvement – – 3.177** – – – – –
(1.232)
Trust – – – .400* – – – –
(.169)
Distrust in government – – – – .950 – – –
(.337)
Income – – – – – 2.311
(1.916)
Personal engagement attitude – – – – – – .663*** –
(.072)
Self-transcendence values – – – – – – – .562***
(.056)
Interactions
Female X GDP 1.282*** – – – – – – –
(.080)
College X GDP – 1.061 – – – – – –
(.113)
Association involvement X GDP – – .916* – – – – –
(.036)
Trust X GDP – – – 1.125** – – – –
(.048)
Distrust in government X GDP – – – – 1.019 – – –
(.036)
(continued)Ethical Consumerism in Global Perspective319
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Table 3. Random Slope Models with Cross-Level Interactions (continued)
Model 5
(Female)
Model 6
(College)
Model 7
(Assoc)
Model 8
(Trust)
Model 9
(Distrust)
Model 10
(Income)
Model 11
(Attitude)
Model 12
(Values)
Income X GDP – – – – – .951 –
(.079)
Personal engagement attitude X GDP– – – – – – 1.045*** –
(.011)
Self-transcendence values X GDP – – – – – – – 1.065***
(.044)
Main effects: country level
GDP per capita (log) 2.624*** 2.678*** 3.031*** 2.124*** 2.536*** 2.784*** 1.629* .870
(.407) (.452) (.524) (.364) (.500) (.480) (.349) (.225)
Postmaterialists (%) 1.040*** 1.054*** 1.048*** 1.051*** 1.054*** 1.053*** 1.057*** 1.061***
(.011) (.011) (.011) (.011) (.012) (.010) (.012) (.012)
Intercept (odds ratios) .000*** .000*** .000*** .000*** .000*** .000*** .000*** .017
(.000) (.000) (.000) (.000) (.000) (.000) (.000) (.043)
Slope variation .013 .086 .009 .006 .007 .041 .000 .000
(.009) (.034) (.004) (.005) (.003) (.018) (.000) (.000)
Notes: Standard errors in parentheses; results presented in odds ratios.Final N for all models ¼ 37,408.
Source: 2004 ISSP citizenship module.
*p < .05 **p < .01 ***p < .001 (two-tailed tests)
320 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

R O B U S T N E S S C H E C K S
Because association involvement was the only variable to exhibit a significant,negative interaction
with GDP,additionalanalyses were performed to determine the robustness ofthat finding.These
include controlling for selectivity into associations,coding association involvement in different way
and operationalizing the country-level economic context in a different way.
Because the measure of association involvement is a composite measure of involve
ferent types of associations it is possible that the finding for association involvement is
tion into types of associations.In other words,it may be that people in low-affluence countries a
more often members of associations that are predisposed to be associated with ethicaconsumerism
than people in high-affluence countries.To controlfor this,country-levelmeasures ofthe share of
membership in each type ofassociation by country were included one by one in models with
originalassociation involvement/GDP interaction.The originalfinding held up to the inclusion of
each new country-levelmeasure,as wellas to severalmeasures that summed the shares of types o
associations, share of political party membership plus union membership, for example
It is also possible that the finding for association involvement simply reflects greater
politics. In other words, it may be that politically interested people are more likely to jo
in the first place and that the finding for association involvement reflects that rather th
ess that occurs as a result of association membership.To address this,political interest was included
in models in two different ways: (1) in an interaction with GDP,and (2) as an individual-level varia-
ble in a modelwith the association involvement/GDP interaction.The interaction between interest
and GDP was marginally significant (p < .10) and positive, suggesting that political int
powerfulpredictor of ethicalconsumerism in high-affluence countries.When interest is included as
an individual-levelcontrolin the model,the negative cross-levelinteraction between association
involvement and GDP remains unchanged (p < .05). These results show that the assoc
mentinteraction finding doesnot reflectgreaterpreexisting politicalinterestamong association
participants.
Interactions were also run for involvement in each type of association on its own to
the original finding was a result of the cumulative effects of being involved in many as
of the five interactions were significantat the p <.10 level(membership in politicalparties and
churches [p ¼ .051]), one was significant at the p < .05 level (membership in sports o
ations),and one was significantatthe p < .01 level(membership in “other” associations).Only
membership in trade unions displayed a non-significant interaction with GDP.Taken together,these
analyses provide further evidence that membership in associations,even in a single type of associa-
tion, is more predictive of ethical consumerism as affluence decreases.
As a finalcheck on robustness,the country-leveleconomic context and association involvemen
were both operationalized in different ways.The economic context was specified as a proxy mea
of the levelof development of ethicalconsumerist industries across country.Countries were given a
1 or 0 along three dimensions: whether they are a member of the Fair Trade Labeling
a member of the Global Ecolabelling Network,and whether the country had a developed market
organic foods in 2003.The finalmeasure was the sum of those indicators.Association involvement
was operationalized including people who belong but don’t actively participate in asso
with people who belong and actively participate (coded together as 1 in a binary meas
weighted measure,following Neilson and Paxton (2010),that gives greater weight to membership i
more explicitly political associations. Results for key interactions remained negative an
D I S C U S S I O N A N D C O N C L U S I O N
By including numerous non-Western countries,this study provides the mostexpansive investiga-
tion ofthe determinants ofethicalconsumerism ofany study to date.The findings confirm the
importance ofthe established individual-and country-levelpredictorsof ethicalconsumerism,
Ethical Consumerism in Global Perspective321
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

but thosegeneralpatternsmasksomeimportantdifferenceshighlightedby the cross-level
interactions.
The low-costhypothesis,which states thatlow-cost(high-affluence) contexts allow individual-
levelattitudes and values to influence behavior,is supported by the positive cross-levelinteractions
between the attitude that engaging in politics is important and self-transcendence valu
Combined with Koos’s (2012) finding for self-transcendence values that mirrors those o
it seems as though the low-cost hypothesis applies well to ethical consumerism in cross
spective.For individual-levelfactorsotherthan valuesand attitudes,however,the story isless
consistent.
Three ofthe six additionalvariables—lack oftrustin governmentinstitutions,education,and
income—exhibited uniform effects across the range ofaffluence.The findings foreducation and
income support the generalizability ofthe idea that ethicalconsumerism is an activity ofthe privi-
leged. Regardless of country-level affluence, higher levels of education and income are
a higher likelihood ofethicalconsumption atthe individuallevel.This suggests thatthe standard
resource modelof politicalparticipation (Brady etal.1995) applies wellto ethicalconsumerism,
even when inclusion of non-Western countries increases the country-levelheterogeneity in analyses.
This does not negate the potentialfor ethicalconsumerism to expand politicalconsciousness of the
marginalized,but it does suggest that,at least for boycotting and buycotting,participatory inequality
is, indeed, a problem (Bartley et al. 2015).
The finding forlack oftrustin governmentinstitutions similarly supports the generalizability
of the idea that dissatisfaction or a lack of trust in conventionalpoliticalinstitutions leads to higher
rates ofnon-electoralparticipation,an idea that boasts a rich research tradition in politicalscience
(Braun and Hutter 2014; Stolle and Micheletti 2013).Ethical consumerism appears to be motivated
globally,by a belief that conventional political institutions cannot be trusted to deliver de
change.
Two of the additionalvariables—gender and generaltrust—displayed a positive significant inter-
action with country-level affluence. As discussed in the results section, the finding for g
that the consistent pattern in previous work that women ethically consume at higher ra
is likely an artifact of the homogeneity of the countries used in analyses.The finding for general trust
is discussed below.
Only one variable—association involvement—produced a significantnegative interaction with
GDP.In other words,association involvementis the only individual-levelvariable whose effectis
stronger in low-affluence countries than in high-affluence countries. This is especially in
sidering the positive interaction between generaltrust,another type ofsocialcapital,and country-
levelaffluence.It suggests thatone ofNeilson and Paxton’s (2010) centralfindings,thathigher
individual-level social capital is associated with higher likelihood of ethical consumption
on the type of socialcapitalanalyzed,country-levelaffluence,and the interaction between the two.
Country-specific logitmodels,notdisplayed here,revealthatgeneraltrustis positively associated
with ethicalconsumerism in 12 of 16 Western countries (75 percent),but in only 4 of 21 countries
in Eastern Europe, Southern Europe, Asia, Latin America, and Israel (19 percent).
One explanation for these divergent findings is that general trust lacks the motivatio
association involvement has.Associations provide people with actionable intelligence regardin
chasing decisions that may encourage ethical consumerism.They also motivate behavior by exposing
people to norms and sanctions for norm violation. These social functions may be espec
in low-affluence countries because ofthe greater structuralconstraints on ethicalconsumerism.In
this way,generaltrust is a “weak” type of socialcapital,whereas association membership is “strong”
in motivating behavior;strong enough,perhaps,to overcome the structuralbarriers to ethicalcon-
sumerism in low-affluence countries.
For many consumers in the wealthy countries of the industrialized West,ethical consumption is a
routine part of shopping. Advertising for fair trade, organic, and eco-labeled products re
322 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

the store and on their couches,often as individuals.In short,Western countries are characterized by
political,cultural,and economic opportunity structures that are more favorable to ethicalconsumer-
ism than are those in most non-Western countries. In such a context the “strong” push
may notbe needed as people have the requisite information and motivation to actregardless of
involvement in associations.When people live in countries with plentifulsupplies of labeled goods,
high aggregate rates of values such as postmaterialism,and political institutions that encourage activ
citizen participation in political issues than general trust is enough to motivate the beh
Although association involvementdoesnot fit neatly into existing accountsof the low-cost
hypothesis (which focus on attitudes and values),this discussion has implications for it.From one
vantage point,this supports the low-cost hypothesis as it is evidence that when externalconditions
render the costs ofa behavior high some extra-individualpush—such as the socialmechanisms at
work in associations—is needed to overcome the structuralbarriers.From a different perspective,it
could be evidence that individual-level values, such as those people may be exposed t
tions,can be important in high-cost contexts,even ifthey need to be bolstered by the aforemen-
tioned social mechanisms.Investigating the role of social mechanisms on behavior in low- an
cost situations would be an interesting direction for future research.
This study suggests additional ways that future research can improve our understan
consumerism.First,the differentialeffects ofthe individual-levelpredictors,including association
involvement,on ethicalconsumerism could be furtherinterrogated by increasing the numberof
countries included in analyses.Recentresearch has shown thatmaximum likelihood models with
cross-levelinteractionscan producebiased estimateswhen thenumberof countriesis small
(Stegmueller 2013).Although the number of countries used here is well within the acceptab
increasing the amountof country-levelheterogeneity even further mightyield additionalinsights.
Perhaps even more importantly,the deeper meanings ofethicalconsumerism across country could
be investigated by performing in-depth case studies within one or a few countries.Especially interest-
ing would be qualitative studies of consumers in low-affluence countries.Non-trivialpercentages of
consumers in such countries report ethical consumption, but what does that look like i
Second,Dietlind Stolle,Marc Hooghe,and Michele Micheletti(2005) were right to callfor the
incorporation of more sophisticated measures of ethical consumerism in large-scale suand that
callis seconded here.Untilsurvey items are developed that allow for meaningfuldistinctions to be
made between boycotting, buycotting, and even other types of ethical consumerism s
motivated brand rejection (Sandıkcıand Ekici2009),ethicalinvesting (Vogel2006),or “discursive
political consumerism” (Stolle and Micheletti 2013),we willlack sophisticated understanding of the
potentialdifferences between those behaviors.For instance,it is possible that the findings for the
resource variables in this study are driven,in part,by the resource-dependent nature ofbuycotting
given that higher percentages of people buycott than boycott (Koos 2012; Yates 2011Friedman’s
(1999) work suggests that the measures we develop should take into account more th
purchase/negative rejection dichotomy that currently animates discussions of the diffe
buycotting and boycotting. Sustained collective buycotts may look much like sustained
cotts in the same way as everyday,one-off decisions to reject or purchase an item may be moti
by similar factors.
Third,the 2004 ISSP is an exemplary data set,butit is notespecially suited to the study of
individual-levelattitudes and values.More sophisticated measures,and measures more specifically
related to ethicalconsumerism,such asattitudestoward fairtrade (Andorfer2013),should be
included in future cross-nationalstudies ofethicalconsumerism.Lastly,cross-nationalstudies of
political participation,including of ethical consumerism,need to seriously consider interaction effect
moving forward.As more and more non-Western data continues to be made available gen
findings based on European data is no longer tenable.It is clear that there are interactions betwee
country-leveland individual-levelpredictorsand the possibility ofanalyzing countrieswith very
Ethical Consumerism in Global Perspective323
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

different political and economic institutions,cultural traditions,and levels of social capital means the
generalizability of findings must now be subject to empirical testing.
Ethicalconsumerism is an important modality ofpoliticalengagement,and it is motivated by a
similar set of individual- and country-level factors as other types of participation. Ethica
may be especially important globally as it has the potential to incorporate into politics t
ized within conventional political processes and to stoke political consciousness among
powerless consumers outside of markets in the Western world. Insofar as association in
powerful predictor of ethical consumerism in low-affluence countries,and in many other ways,it has
a role to play in global politics moving forward.
R E F E R E N C E S
Adams,Matthew and Jayne Raisborough.2010.“Making a Difference:EthicalConsumption and the Everyday.”
The British Journal of Sociology 61:256-74.
Almond, Gabriel A. and Sidney Verba. [1963] 2015. The Civic Culture: Political Attitudes and Democr
Princeton, NJ: Princeton University Press.
Andorfer,Veronika A.2013.“EthicalConsumption in Germany.A Cross-SectionalAnalysis of Determinants of Fair
Trade Consumption (2000–2010).” Zeitschrift fu¨r Soziologie 42:424-43.
Andorfer,Veronika A.and UlfLiebe.2013.“Consumer Behavior in MoralMarkets.On the Relevance ofIdentity,
JusticeBeliefs,SocialNorms,Status,and Trust in EthicalConsumption.”EuropeanSociologicalReview
29(6):1251-65.
Andrews,Kenneth T.,MarshallGanz,Matthew Baggetta,Hahrie Han,and Chaeyoon Lim.2010.“Leadership,
Membership, and Voice: Civic Associations That Work.” American Journal of Sociology 115:1191-24
Baggetta, Matthew. 2009. “Civic Opportunities in Associations: Interpersonal Interaction, Governance
Institutional Relationships.” Social Forces 88:175-99.
Barnett, Clive, Paul Cloke, Nick Clarke, and Alice Malpass. 2011. Globalizing Responsibility: The Politi
Ethical Consumption. West Sussex, UK: Wiley-Blackwell
Barrett,Martyn and Ian Brunton-Smith.2014.“Politicaland Civic Engagementand Participation:Towardsan
Integrative Perspective.” Journal of Civil Society 10:5-28.
Bartley,Tim.2007.“Institutional Emergence in an Era of Globalization: The Rise of Transnational Private R
of Labor and Environmental Conditions.” American Journal of Sociology 113:297-351.
Bartley,Tim,Sebastian Koos,Gustavo Setrini,Hiram Satel,and Nik Summers.2015.Looking Behind the Label: Global
Industries and the Conscientious Consumer. Bloomington: Indiana University Press.
Bevir,Mark and Frank Trentmann.2007.Governance,Consumers and Citizens: Agency and Resistance in Contem
Politics. New York: Palgrave Macmillan
Bostro¨m,Magnus and MikaelKlintman.2008.Eco-Standards,Product Labelling and Green Consumerism.Basingstoke,
UK: Palgrave Macmillan
Brady,Henry E.,Sidney Verba,and Kay Lehman Schlozman.1995.“Beyond SES:A Resource Modelof Political
Participation.” American Political Science Review 89:271-94.
Braun,Daniela and Swen Hutter.2014.“PoliticalTrust,Extra-RepresentationalParticipation,and the Openness of
PoliticalSystems.”InternationalPoliticalScienceReview.RetrievedMarch 15, 2015 (DOI:10.1177/
0192512114559108).
Breen,Timothy Hall.2004.The Marketplace of Revolution: How Consumer Politics Shaped American IndepeNew
York: Oxford University Press.
Clark,Burton R.1986.The Higher Education System:Academic Organization in Cross-NationalPerspective.Berkeley:
University of California Press.
Clarke, Nick. 2008. “From Ethical Consumerism to Political Consumption.” Geography Compass 2:187
Clarke,Nick, CliveBarnett,PaulCloke,and AliceMalpass.2007.“ThePoliticalRationalitiesof Fair-Trade
Consumption in the United Kingdom.” Politics & Society 35:583-607.
Cohen,Lizabeth.2004.A Consumers’Republic:The Politicsof MassConsumption in PostwarAmerica.New York:
Random House.
Coleman, James S. 1990. Foundations of Social Theory. New York: Belknap Press.
Connolly,John and Andrea Prothero.2008.“Green Consumption Life-Politics,Risk and Contradictions.” Journalof
Consumer Culture 8:117-45.
Copeland,Lauren.2014.“Conceptualizing PoliticalConsumerism: How Citizenship Norms Differentiate Boycottin
from Buycotting.” Political Studies 62:172-86.
324 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Dauvergne, Peter and Jane Lister. 2013. Eco-Business: A Big-Brand Takeover of Sustainability. Camb
Press.
Davies,Kate.2001.“A MoralPurchase:Femininity,Commerce and Abolition,1788–1792.” Pp.133-61 in Women,
Writing,and thePublicSphere:1700-1830,edited by E.Eger,C. Grant,C. O. Gallchoir,and P.Warburton.
Cambridge, UK: Cambridge University Press.
Davis, Deborah. 2000. The Consumer Revolution in Urban China. Berkeley: University of California P
——. 2006.“Urban Chinese Homeownersas Consumer-Citizens.”Pp. 281-301 in TheAmbivalentConsumer:
Questioning Consumption in East Asie and the West, edited by S. M. Garon and P. L. Maclachlan.
University Press
Derksen, Linda and John Gartrell. 1993. “The Social Context of Recycling.” American Sociological Re
Diekmann,Andreas and Peter Preisendo¨rfer.2003.“Green and Greenback the BehavioralEffects ofEnvironmental
Attitudes in Low-Cost and High-Cost Situations.” Rationality and Society 15:441-72.
Djelic, Marie-Laure and Kerstin Sahlin-Andersson. 2006. Transnational Governance: Institutional Dyn
Cambridge, UK: Cambridge University Press.
Dreher,Axel.2006.“DoesGlobalization AffectGrowth? Evidence from a New Index ofGlobalization.” Applied
Economics 38:1091-110.
Ecolabel Index. 2014. “Ecolabel Index Home Page.” Retrieved June 15, 2014 (www.ecolabelindex.co
Fair Trade Foundation.2014.“Facts,Figures & Resources.” Retrieved June 15,2014 (www.fairtrade.org.uk/press_
office/facts_figures.aspx).
Fair Trade USA.2012.“Fair Trade USA Almanac 2012.” Retrieved September 1,2014 (http://fairtradeusa.org/sites/
default/files/2012_Fair-Trade-USA_Almanac.pdf).
Ferrer,Mariona and Marta Fraile.2006.“Exploring the SocialDeterminants ofPoliticalConsumerism in Western
Europe.” Working Paper,Universidad Autonoma de Madrid.Departamento de Ciencia Polıtica y Relaciones
Internacionales, Madrid, Spain.
Forno,Francesca and LuigiCeccarini.2006.“From the Streetto the Shops:The Rise ofNew Forms ofPolitical
Actions in Italy.” South European Society and Politics 11:197-222.
Frank, Dana. 2000. Buy American: The Untold Story of Economic Nationalism. Boston, MA: Beacon P
Franklin,Mark N.2004.Voter Turnout and the Dynamics of Electoral Competition in Established Democra
Cambridge, UK: Cambridge University Press.
Friedman, Monroe. 1996. “A Positive Approach to Organized Consumer Action: The ‘Buycott’ as an A
Boycott.” Journal of Consumer Policy 19:439-51.
——. 1999. Consumer Boycotts: Effecting Change through the Marketplace and the Media. New York
Fukuyama, Francis. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York: Free Pr
Global Ecolabeling Network (GEN).2014.Global Ecolabelling Network Home Page.Retrieved June 15,2014 (www.
globalecolabelling.net/members_associates/map/index.htm).
Gereffi,Gary and MiguelKorzeniewicz.1994.Commodity Chainsand GlobalCapitalism.Westport,CT: Praeger
Publishers.
Glickman,Lawrence B.2004.“‘Buy for the Sake of the Slave’: Abolitionism and the Origins of American Co
Activism.” American Quarterly 56:889-912.
Goodman,MichaelK. 2010.“The Mirror ofConsumption:Celebritization,DevelopmentalConsumption and the
Shifting Cultural Politics of Fair Trade.” Geoforum 41:104-16.
Granovetter,Mark.1985.“Economic Action and Social Structure: The Problem of Embeddedness.” Americ
of Sociology 91:481-510.
Grasseni,Cristina.2003.“Packaging Skills:Calibrating Cheese to the GlobalMarket.” Pp.259-87 in Commodifying
Everything, edited by S. Strasser. New York: Routledge.
Greenberg,Cheryl.1999.“Don’t Buy Where You Can’t Work.” Pp.241-75 in Consumer Society in American History
A Reader, edited by L. B. Glickman. Ithaca, NY: Cornell University Press.
Guagnano,Gregory A.,PaulC. Stern,and Thomas Dietz.1995.“Influences on Attitude-Behavior Relationships a
Natural Experiment with Curbside Recycling.” Environment and Behavior 27:699-718.
Guthman,Julie.2007.“Can’t Stomach It: How MichaelPollan et al.Made Me Want to Eat Cheetos.” Gastronomica:
The Journal of Critical Food Studies 7:75-79.
——.2008. “‘If They Only Knew’: Color Blindness and Universalism in California Alternative Food Inst
Professional Geographer 60:387-97.
Hainmueller,Jensand MichaelHiscox.2015.“The Socially ConsciousConsumer? Field ExperimentalTestsof
Consumer Support for Fair Labor Standards.” SocialScience Research Network.Retrieved June 6,2016 (http://pa
pers.ssrn.com/sol3/papers.cfm?abstract_id¼2062435).
Ethical Consumerism in Global Perspective325
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Hainmueller, Jens, Michael Hiscox, and Sandra Sequeira. 2015. “Consumer Demand for the Fair Trad
from a Field Experiment.” Review of Economics and Statistics 97:242-56.
Hilton,Matthew.2002.“The Female Consumer and the Politics of Consumption in Twentieth-Century Brita
Historical Journal 45:103-28.
——. 2003.Consumerism in Twentieth-CenturyBritain:TheSearch fora HistoricalMovement.Cambridge,UK:
Cambridge University Press.
Hobson,Kersty.2004.“Researching‘SustainableConsumption’in Asia-PacificCities.”AsiaPacificViewpoint
45:279-88.
Huckfeldt, R. Robert. 1979. “Political Participation and the Neighborhood Social Context.” American J
Science 23:579-92.
Hunter,Lori M., Alison Hatch,and Aaron Johnson.2004.“Cross-NationalGenderVariation in Environmental
Behaviors.” Social Science Quarterly 85:677-94.
Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Society. Princeton, NJ: Princeton Univers
——.1997.Modernization and Postmodernization: Cultural,Economic,and PoliticalChange in 43 Societies.Cambridge,
UK: Cambridge Univ Press.
InternationalSocialSurvey Program (ISSP).2004."InternationalSocialSurvey Program Citizenship Module.”
RetrievedJuly 1, 2012 (http://zacat.gesis.org/webview/index.jsp?object¼http://zacat.gesis.org/obj/f
ZA3950)
Iversen,Torben and FrancesRosenbluth.2006.“The PoliticalEconomy ofGender:Explaining Cross-National
Variation in the Gender Division ofLabor and the Gender Voting Gap.” American Journalof PoliticalScience
50:1-19.
Jacobsen,Eivind and Arne Dulsrud.2007. “Will Consumers Save the World? The Framing of Political Consume
Journal of Agricultural and Environmental Ethics 20:469-82.
Jepperson,Ronald L.2002.“PoliticalModernities:Disentangling Two Underlying Dimensionsof Institutional
Differentiation.” Sociological Theory 20:61-85.
Johnston,Jose´e.2008.“The Citizen-Consumer Hybrid: Ideological Tensions and the Case of Whole Foods M
Theory and Society 37:229-70.
Johnston, Jose´e, Michelle Szabo, and Alexandra Rodney. 2011. “Good Food, Good People: Understan
Repertoire of Ethical Eating.” Journal of Consumer Culture 11:293-318.
Johnston, Jose´e and Shyon Baumann. 2010. Foodies: Democracy and Distinction in the Gourmet Foo
Routledge.
Kam,Cindy D.and CarlL. Palmer.2008.“Reconsidering the Effects ofEducation on PoliticalParticipation.” The
Journal of Politics 70:612-31.
Klein,Jakob.2009.“Creating Ethical Food Consumers? Promoting Organic Foods in Urban Southwest Chin
Anthropology 17:74-89.
Koos,Sebastian.2011.“Varieties of Environmental Labelling,Market Structures,and Sustainable Consumption across
Europe:A Comparative Analysis of Organizationaland Market Supply Determinants ofEnvironmental-Labelled
Goods.” Journal of Consumer Policy 34:127-51.
——. 2012.“WhatDrives PoliticalConsumption in Europe? A MultilevelAnalysis on IndividualCharacteristics,
Opportunity Structures, and Globalization.” Acta Sociologica 55:37-57.
Kroneberg,Clemens,Isolde Heintze,and Guido Mehlkop.2010.“The Interplay ofMoralNorms and Instrumental
Incentives in Crime Causation.” Criminology 48:259-94.
Leighley,Jan E.1995.“Attitudes,Opportunities,and Incentives:A Field Essay on PoliticalParticipation.” Political
Research Quarterly 48:181-209.
Liebe,Ulf and Peter Preisendo¨rfer.2010.“Rational Choice Theory and the Environment: Variants,Applications,and
New Trends.” Pp.140-57 in EnvironmentalSociology: European Perspectives and Interdisciplinary Challengeedited
by M. Gross and H. Heinrichs. Dordrecht, Netherlands: Springer Science
Malpass,Alice,Clive Barnett,Nick Clarke,and PaulCloke.2007.“Problematizing Choice: Responsible Consumers
and ScepticalCitizens.” Pp.231-56 in Governance,Consumers,and Citizens: Agency and Resistance in Contempor
Politics, edited by M. Bevir and F. Trentmann. Basingstoke, UK: Palgrave MacMillan.
Maniates, Michael F. 2001. “Individualization: Plant a Tree, Buy a Bike, Save the World?” Global Envir
1:31-52.
Maniates, Michael and John M. Meyer. 2010. "The Environmental Politics of Sacrifice." Cambridge, MA
Marien,Sofie,Marc Hooghe,and Ellen Quintelier.2010.“Inequalities in Non-Institutionalised Forms ofPolitical
Participation: A Multi-Level Analysis of 25 Countries.” Political Studies 58:187-213.
326 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Marshall,Monty G.,Keith Jaggers,and Ted Robert Gurr.2011.Polity IV Project: PoliticalRegime Characteristics and
Transitions, 1800-2010. College Park: University of Maryland
McAdam,Doug,John D.McCarthy,and Mayer N.Zald.1996.Comparative Perspectives on Social Movements: Po
Opportunities, Mobilizing Structures, and Cultural Framings. Cambridge, UK: Cambridge Universit
Micheletti,Michele.2003.PoliticalVirtue and Shopping:Inviduals,Consumerism,and Collective Action.New York:
Palgrave Macmillan.
Micheletti,Michele and Dietlind Stolle.2005.“Swedish PoliticalConsumers:Who They Are and Why They
Use the Marketas an Arenafor Politics.”Pp. 26-29 in PoliticalConsumerism:Its Motivations,Power,and
Conditions in the Nordic Countries and Elsewhere,edited by M.Bostro¨m,A.Follesdal,M. Klintman,M. Micheletti,
and M. P. Sorensen. Copenhagen, Denmark: Norden.
Mutz,Diana C.2002.“Cross-Cutting SocialNetworks:Testing Democratic Theory in Practice.” American Politic
Science Review 96:111-26.
Nannestad,Peter.2008.“What Have We Learned About Generalized Trust,If Anything?” AnnualReview ofPolitical
Science 11:413-36.
Neilson,Lisa A.2010.“Boycott or Buycott? Understanding PoliticalConsumerism.” JournalofConsumer Behaviour
9:214-27.
Neilson,Lisa A.and Pamela Paxton.2010.“SocialCapitaland PoliticalConsumerism: A MultilevelAnalysis.” Social
Problems 57:5-24.
Nelson,Laura C.2000.Measured Excess:Status,Gender,and ConsumerNationalism in South Korea.New York:
Columbia University Press.
Organic Trade Association (OTA).2011.“OTA Industry Statistics and Projected Growth.” Retrieved June 15,2014
(www.ota.com/organic/mt/business.html).
——. 2012. “OTA Press Report 2012.” Retrieved June 15, 2014 (www.organicnewsroom.com/2012/04
driven_organic_mark.html).
Putnam,RobertD. 1994.MakingDemocracyWork:CivicTraditionsin Modern Italy.Princeton,NJ: Princeton
University Press.
Reich,RobertB. 2008.Supercapitalism:The Transformation ofBusiness,Democracy,and Everyday Life.New York:
Random House.
Rosenstone,Steven and John M Hansen.1993.Mobilization,Participation,and Democracy in America.New York:
Macmillan.
Ro¨ssel, Jo¨rg. 2008. “Conditions for the Explanatory Power of Life Styles.” European Sociological Re
Sandıkcı,€Ozlem and AhmetEkici.2009.“Politically Motivated Brand Rejection.”Journalof BusinessResearch
62:208-17.
Sassatelli,Roberta.2006.“Virtue,Responsibility,and Consumer Choice.” Pp.219-50 in Consuming Cultures,Global
Perspectives: HistoricalTrajectories,TransnationalExchanges,edited by J.Brewer and F.Trentmann.Oxford,UK:
Berg Publishers.
——. 2007. Consumer Culture: History, Theory, and Politics. London, UK: Sage.
Sassatelli, Roberta and Alan Scott.2001. “Novel Food, New Markets, and Trust Regimes: Responses to the Er
Consumers’ Confidence in Austria, Italy and the UK.” European Societies 3:213-44.
Sassatelli,Roberta and Federica Davolio.2010.“Consumption,Pleasure,and Politics Slow Food and the Politico-
Aesthetic Problematization of Food.” Journal of Consumer Culture 10:202-32.
Schneiberg,Marc and Tim Bartley.2008.“Organizations,Regulation,and Economic Behavior: Regulatory Dynamics
and Forms from the Nineteenth to Twenty-First Century.” Annual Review of Law and Social Scienc
Schofer,Evan and Marion Fourcade-Gourinchas.2001.“The StructuralContexts ofCivic Engagement:Voluntary
Association Membership in Comparative Perspective.” American Sociological Review 66:806-28.
Schons,Laura Marie,Mario Rese, and Jan Wieseke. 2010.Rational Choice under Pressure: An Empirical Investigati
of Behavior in Low-Costand High-CostSituations.Retrieved June 6,2016 (www-brs.ub.ruhr-uni-bochum.de/
netahtml/HSS/Diss/SchonsLauraMarie/diss.pdf#page¼122).
Schudson,Michael.2007.“Citizens,Consumers,and the Good Society.” The ANNALS ofthe American Academy of
Political and Social Science 611:236-49.
Scott,James C.[1985] 2008.Weaponsof the Weak:Everyday FormsofPeasantResistance.New Haven,CT: Yale
University Press.
Seidman,Gay.2007.Beyond the Boycott: Labor Rights,Human Rights,and TransnationalActivism.New York: Russell
Sage Foundation.
Shaw,Deirdre,Emma Grehan,Edward Shiu,Louise Hassan,and Jennifer Thomson.2005.“An Exploration of Values
in Ethical Consumer Decision Making.” Journal of Consumer Behaviour 4:185-200.
Ethical Consumerism in Global Perspective327
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from

Shaw,Deirdre,Terry Newholm,and Roger Dickinson.2006.“Consumption as Voting: An Exploration of Consumer
Empowerment.” European Journal of Marketing 40:1049-67.
Smith,Toby Maureen.1998.The Myth ofGreen Marketing:Tending Our Goats atthe Edge ofApocalypse.Toronto,
Canada: University of Toronto Press.
Steering Committee of the State-of-Knowledge Assessment of Standards and Certification. 2012. Tow
The Roles and Limitations of Certification. Washington, DC: RESOLVE, Inc.
Stegmueller, Daniel. 2013. “How Many Countries for Multilevel Modeling? A Comparison of Frequenti
Approaches.” American Journal of Political Science 57:748-61.
Stolle, Dietlind and Marc Hooghe. 2011. “Shifting Inequalities: Patterns of Exclusion and Inclusion in
of Political Participation.” European Societies 13:119-42.
Stolle, Dietlind, Marc Hooghe, and Michele Micheletti. 2005. “Politics in the Supermarket: Political Co
Form of Political Participation.” International Political Science Review 26:245-69.
Stolle,Dietlind and Michele Micheletti.2006.“The Gender Gap Reversed:PoliticalConsumerism as a Women-
Friendly Form of Civic and Political Engagement.” Pp. 45-72 in Gender and Social Capital, edited b
E. Gidengil. New York: Routledge.
——. 2013. Political Consumerism: Global Responsibility in Action. Cambridge, UK: Cambridge Univer
Strømsnes,Kristin.2005.“Political Consumption in Norway: Who,Why—and Does It Have Any Effect.” Pp.165-81
in Political Consumerism: Its Motivations,Power,and Conditions in the Nordic Countries and Elsewhere,edited by M.
Bostro¨m, A. Follesdal, M. Klintman, M. Micheletti, and M. P. Sorensen. Copenhagen, Denmark: No
Sunderer, Georg and Jo¨rg Ro¨ssel. 2012. “Morality or Economic Interest? The Impact of Moral Motive
Factors on the Purchase of Fair Trade Groceries.” International Journal of Consumer Studies 36:24
Szasz,Andrew.2007.Shopping Our Way to Safety:How We Changed from Protecting the Environmentto Protecting
Ourselves. Minneapolis: University of Minnesota Press.
Thøgersen,John.2010.“Country Differences in Sustainable Consumption:The Case ofOrganic Food.” Journalof
Macromarketing 30:171-85.
Tobiasen,Mette.2005.“PoliticalConsumerism in Denmark.” Pp.113-43 in PoliticalConsumerism:ItsMotivations,
Power,and Conditions in the Nordic Countries and Elsewhere,edited by M.Bostro¨m,A. Follesdal,M. Klintman,M.
Micheletti, and M. P. Sorensen. Copenhagen, Denmark: Norden.
Verba,Sidney,Kay Lehman Schlozman,and Henry E Brady.1995.Voice and Equality: Civic Voluntarism in American
Politics. Cambridge, UK: Cambridge University Press.
Verba,Sidney and Norman H Nie.1972.Participation in America: PoliticalDemocracy and SocialEquality.Chicago:
University of Chicago Press.
Vogel,David.2006. The Market for Virtue: The Potential and Limits of Corporate Social Responsibility.Washington,DC:
Brookings Institution Press.
Vrablıkova,Katerina.2013.“How Context Matters? Mobilization,PoliticalOpportunity Structures,and Nonelectoral
PoliticalParticipation in Old and New Democracies.” Comparative PoliticalStudies.Retrieved March 15,2015
(DOI: 10.1177/0010414013488538).
Wahlstro¨m, Mattias and Abby Peterson. 2006. “Between the State and the Market: Expanding the C
Opportunity Structure’.” Acta Sociologica 49:363-77.
Warde, Alan. 2005. “Consumption and Theories of Practice.” Journal of Consumer Culture 5:131-53.
Willis,Margaret M.and Juliet B.Schor.2012.“Does Changing a Light Bulb Lead to Changing the World? Politic
Action and the Conscious Consumer.” TheANNALS ofthe American Academy ofPoliticaland SocialScience
644:160-90.
Yan,Yungxiang.2000.“Of Hamburger and SocialSpace:Consuming McDonald’s in Beijing.” Pp.201-25 in The
Consumer Revolution in Urban China, edited by D. Davis. Berkeley: University of California Press.
Yates, Luke S. 2011. “Critical Consumption: Boycotting and Buycotting in Europe.” European Societie13:191-217.
Zelizer, Viviana A. 2009. The Purchase of Intimacy. Princeton, NJ: Princeton University Press.
328 Summers
by guest on September 6, 2016http://socpro.oxfordjournals.org/Downloaded from