-
A Political Mediation Model of Corporate Response to Social
Movement Activism1
Brayden G King2
Brigham Young University
Word count: 12,433
Revised: September 5, 2007
Forthcoming in Administrative Science Quarterly
1 The author would like to acknowledge the valuable feedback
received from Tim Bartley, Brandon Lee, Tina
Fetner, Peter Madsen, and Marie Cornwall. 2 Brayden King,
Department of Sociology, 2045 JFSB, Brigham Young University,
Provo, UT 84602,
[email protected].
-
2
A Political Mediation Model of Corporate Response to Social
Movement Activism
ABSTRACT
This paper uses a theory of social movement outcomes, the
political mediation model, to explain
why certain corporations targeted by boycotts were more likely
to concede to boycotters
demands. We should expect that boycotts threaten tangible and
intangible resources held by
corporate targets, that these threats are transmitted indirectly
through media coverage of the
boycotts, that past performance declines create opportunities
for movement influence, and that
the level of threat posed by a boycott generates more influence
when targeted against
corporations that recently experienced performance declines.
Using a sample of corporate
boycotts reported in major national newspapers between 1990 and
2005, the results provide
support for the political mediation model. Corporate targets of
boycotts were more likely to
concede when the boycott received a great deal of media
attention. The effect of media attention
was amplified when the corporate target previously experienced a
reputational decline.
-
3
Much research at the nexus of social movement analysis and
organizational theory examines the
ways in which secondary stakeholders (i.e., communities,
activist groups) try to influence a
firms practices or policies (Hoffman, 1999; Lounsbury, 2001;
Scully and Segal, 2002;
Lounsbury, Ventresca and Hirsch, 2003; Raeburn, 2004; Davis,
McAdam, Scott, and Zald 2005).
Increasingly, scholars are interested in understanding how
social movements convince
organizational leaders to make changes to their policies or
practices (Berry, 2003; Baron, 2003;
Schurman, 2004; Luders, 2006; Rojas, 2006; 2007). A motivating
question is, without seeking
the state as a mediator, when are social movement activists - a
relatively powerless community
of individuals - able to influence managers to change their
policies or practices?
The question is important given certain expectations established
by other theoretical
perspectives. Many corporate governance scholars, for example,
assert that managers will not or
should not alter policies unless the change favors shareholders
(e.g. Berle and Means, 1932;
Smith, 1998). Organizational sociologists have argued that in
the current era of investor
capitalism shareholders interests have primacy over other
decision-making criteria (Conard,
1988; Useem, 1996; Fligstein, 2001; Khurana, 2002). According to
these perspectives, under the
current corporate governance regime secondary stakeholders
should have little influence over
organizational decision-making.
In contrast, stakeholder theory proposes that various
stakeholder groups have legitimate claims
on corporate governance (e.g. Donaldson and Preston, 1995;
Jones, 1995; Clarkson, 1995). In
addition to shareholders, employees, customers, and communities
want to provide input to
corporate decision-making. Beyond offering a descriptive theory
of the corporation, some
-
4
stakeholder scholarship attempts to distinguish various sources
of stakeholder influence (e.g.
Frooman, 1999). Lacking a direct voice in corporate
decision-making due to insignificant
control of shareholder wealth, secondary stakeholders must find
other means to influence
organizational change. Stakeholder theory, however, has been
criticized for not identifying the
causal linkages whereby stakeholders exert influence (Key,
1999). The stakeholder theory of the
firm, critics claim, is not sufficiently predictive or
explanatory. This criticism calls for the need
to develop conceptual mechanisms to explain when and how
stakeholders should matter.
Understanding the conditions in which stakeholder influence is
possible should be a priority for
organizational scholars interested in stakeholder-initiated
change (e.g. Gargiulo, 1993; Rowley,
1997; Mitchell, Agle, and Wood, 1997; Frooman, 1999).
In a parallel literature, social movement scholars have studied
how activist-initiated change
occurs in the political realm (e.g., Giugni, 1999). Research on
social movement outcomes
focuses on a similar problem: when can activists with little
direct control over legislative
decision-making influence policymakers to initiate change?
Scholars have identified a number of
factors that facilitate social movement influence, including
characteristics of the movement and
characteristics of the movement target (Amenta and Young, 1999;
Burstein and Linton, 2002;
Soule and Olzak, 2004; Giugni, 2004; Soule and King, 2006). In
recent years these insights have
also been applied to the study of social movement influence over
organizations (Raeburn, 2004;
Zald, Morrill and Rao, 2005; Strang and Jung, 2005; Luders,
2006; King and Soule, 2007; Rojas,
2006; 2007).
-
5
Several main themes appear in this literature. First, social
movements can help change political
and organizational agendas by using extrainstitutional tactics
or tactics not commonly
embraced by institutional elites and that, in some way, are
subversive to the target institution
(Gamson, 1990; Cress and Snow, 2000; Rojas, 2006; King, Bentele,
and Soule, 2007; King and
Soule, 2007). Extrainstutional tactics draw attention to
perceived injustices and broaden political
discussion to include voices not typically represented in
legitimate channels of change. Protest,
for example, is an extrainstitutional tactic that highlights the
outsider role of activists and appeals
to third-parties by broadcasting grievances publicly (Lipsky,
1968). Second, exogenous factors
or political opportunities as they are often called predispose
certain movement targets to
change (Jenkins and Perrow, 1977; Kitschelt, 1986; Tarrow, 1994;
Soule et al., 1999; Meyer and
Minkoff, 2004; Soule and King, 2006). These movement opportunity
structures spur activists to
take action by signaling possibilities for movement influence
(Tarrow, 1996; Schurman, 2004;
Raeburn, 2004). Opportunities for activism also predispose
political targets to be open to change
(Soule and King, 2006). A third contribution to the study of
movement outcomes is the political
mediation model developed by Amenta and his colleagues (Amenta,
Carruthers, and Zylan,1992;
Amenta, Dunleavy, and Bernstein, 1994; Amenta and Young, 1999;
Cress and Snow, 2000;
Amenta, 2006). This model proposes that movements are more
influential in some contexts than
others. The more open the target institution is to change, the
more effective movement
mobilization will be. Thus, the effects of extrainstitutional
tactics and opportunities are thought
to be interactive, not additive. Movement influence is moderated
by the institutional proclivity
to change.
-
6
Following Amentas (2006) call to more carefully conceptualize
the contexts in which activist
influence is amplified, this paper extends and refines the
political mediation model by specifying
what favorable conditions for change look like in the corporate
setting. Movement mobilization
should be most effective when targeted against firms that are
already structurally predisposed to
external influence. Extending the political mediation model
requires identifying those structural
opportunities that weaken corporate resistance to change and
facilitate activist influence. With
this end in mind, the analysis in this paper assesses the
factors leading to corporate concessions
to social movement boycotts.
Boycotts are a quintessential tactic of the anti-corporate,
labor, and other social change
movements (Manheim, 2001). Activists often use boycotts because
they lack other means of
influence and because boycotts have proven to be a reliable
tactic to initiate organizational and
social change (Friedman, 1999). Evidence, however, suggests that
boycotts vary in their
effectiveness (Mahoney, 1976; Miller and Sturdivant, 1977;
Zadek, 2001; Vogel, 2005;
Coleman, Lee, and Rubinowitz, 2005). By explaining the
determinants of boycotts ability to
influence executives to change corporate practices and policies,
we improve our understanding
of the conditions wherein activists and other secondary
stakeholders acquire influence over the
corporation.
I propose that boycott effectiveness is determined by boycotters
ability to threaten the corporate
target through media accounts and by movement opportunities
constituted as a past decline in the
organizational targets sales revenue and reputation. While most
research on boycotts focuses on
the ability of boycotters to divert revenue from the corporate
target (e.g. Friedman, 1999), this
-
7
study considers a second possible mechanism whereby boycotts
influence their targets
threatening the companys public image. One of the ways that
activists may threaten
corporations is through their ability to harm the targets
corporate image, thereby damaging the
corporations reputation (Fombrun and Shanley, 1990; Bromley,
1993; Fombrun, 1996; Elsbach
and Kramer, 1996; Rindova and Fombrun, 1999; Deephouse, 2000;
Whetten and Mackey, 2002).
Prior performance declines in sales revenue and reputation make
firms vulnerable to attack by
boycotters. Following the political mediation model, I expect
that the effectiveness of boycott
threat as a way to gain corporate concessions is moderated by
corporate performance declines.
Explaining Managerial Concessions to Boycotters
Boycotts are a widely used movement tactic for activists trying
to persuade corporate targets to
adopt some change in practice or policy (Putnam, 1993; Manheim,
2001; John and Klein, 2003).
By refusing to buy a companys services or products, consumers
express their dissatisfaction
with the target and make political claims about corporate
practices (Friedman, 1985). Boycotts
are deemed successful when the demands of the boycotters are met
by the corporate target.
Typically this means that the corporation makes an explicit and
publicly-recognized effort to
address the concerns of the boycotters, causing an end to the
boycott. For example, in May of
2001 the Toyota Motor Corporation pulled a television
advertisement depicting an African-
American man with a tooth inlaid with a gold Toyota insignia
after the Rainbow Coalition, a
civil rights organization led by Jesse Jackson, accused Toyota
of reinforcing racial stereotypes
and called for a boycott against the company. In addition to
pulling the advertisement, Toyota
announced plans to extend its business relations with minority
clients, banks, and suppliers.
Toyota reportedly increased its minority-spending budget by
thirty-five percent (Greising, 2001).
-
8
By giving in to the demands of the activists and making further
policy corrections, Toyota
clearly offered a concession to the boycotters.
Not all boycotts are equally effective at getting concessions
(Mahoney, 1976; Miller and
Sturdivant, 1977; Zadek, 2001; Vogel, 2005; Coleman, Lee, and
Rubinowitz, 2005). An
empirical examination of corporate concessions to boycotter
demands found that only a quarter
of all publicized boycotts were successful (Friedman, 1985).
Event studies examining boycotts
effects on stock price returns also have mixed results. Three
studies found that boycotts (or
boycott announcements) generated negative stock price returns
(Pruitt and Friedman, 1986;
Pruitt, Wei, and White, 1988; Davidson, Worrell, and El-Jelly,
1995), but another study found
that some boycotts actually led to positive returns (Koku,
Akhigbe, and Springer, 1997). Thus,
the evidence suggests that boycotts vary in their
effectiveness.
A lack of theoretical development has impeded researchers from
forming a coherent explanation
for why some corporate targets are more willing to yield to the
demands of boycotts. Research
has not explained why some boycotts are viewed as more
threatening than others nor has it
focused attention on the ways in which boycott threat affects
changes in the corporate context.
Using insights from social movement and organizational theories,
this analysis helps fill this
theoretical and analytical gap.
The maturing literature on social movement outcomes identifies a
number of factors that shape
movements abilities to achieve their goals (e.g. Burstein and
Linton, 2002; Soule and King,
2006). A number of theorists have pointed to extrainstitutional
tactics, like protest, as a
-
9
mechanism underlying social movement influence (Gamson, 1990;
Cress and Snow, 2000;
Andrews, 2001; McAdam and Su, 2002; Rojas, 2006; King and Soule,
2007; King, Bentele, and
Soule, 2007). This study follows that tradition by focusing on
extrainstitutional tactics as a
means to gain influence when access to legitimate channels of
change is blocked.
The effect of extrainstitutional tactics
One of the defining features of social movements is their use of
extrainstitutional tactics
activities designed to subvert conventional politics and bypass
traditional inputs (McAdam,
1982). Activists may use extrainstitutional tactics, like
boycotts or protests, when they lack
access to legitimate channels of change or when their influence
has been muffled by insider
politics. Lack of access to legitimate avenues may be especially
likely when the target of change
is a business corporation. Corporations have no legal
responsibility to secondary stakeholders;
their primary obligation is to shareholders (e.g. Smith, 1998).
Business corporations also have
the legal mandate to seek profit above all else, thus
legitimating the claim that secondary
stakeholders, like activists, are of no concern when setting
policies (Friedman, 1962).
Thus, it comes as no surprise that activists increasingly use
boycotts or other nonconventional
tactics as means of influence and expression when other avenues
of change have been blocked
(John and Klein, 2003).3 Social movement scholars have found
that the use of extrainstitutional
tactics often leads to movement success (Gamson, 1975; Piven and
Cloward, 1977; Mirowsky
and Ross, 1981; Rojas, 2006; Luders, 2006). Extrainstitutional
tactics work for two main
reasons. First, these tactics may disrupt organizational
routines and impose costs on the target
3 In fact, one may assert that the use of boycotts is prima
facie evidence that insider paths to legitimate
organizational change have been blocked.
-
10
(Luders, 2006). By disrupting routine activities,
extrainstitutional tactics threaten to constrain
the target organization from obtaining critical resources.
Boycotts are an exemplary form of this
tactic, as they threaten to constrain revenue and impose
marketing costs (Garrett, 1987;
Friedman, 1999). Boycotts are successful insofar as they take
potential revenue away from the
firm, thereby putting pressure on managers to conform.
The second way in which extrainstitutional tactics foster
success is through indirect influence by
appealing to broader organizational audiences. This second
mechanism of influence relies on the
boycotts potential to shape public perceptions of the firm,
regardless of the direct financial
losses the boycott might cause. By bypassing legitimate channels
of change, extrainstitutional
tactics publicly broadcast movement grievances and appeal to the
sensibilities of the target
organizations primary audiences (e.g. customers). In the
political realm protests indirectly
influence legislators by shaping public opinion regarding a
specific issue and by changing the
calculus of voters; in the corporate context, activists seek to
shape public perceptions of a firm.
Boycotts and other movement tactics present alternative messages
about a target that threaten its
efforts to build a positive image among constituents (Garrett,
1987; Basdeo, Smith, Grimm,
Rindova, and Derfus, 2006). Business organizations are
particularly susceptible to attacks
against their images because their survival depends on control
over key tangible and intangible
resources that may be threatened by these tactics (Pfeffer and
Salancik, 1978; Barney, 1991).
Damage to their image may devalue their established reputations
(Fombrun, 1996; Whetten and
Mackey, 2002; Deephouse and Carter, 2005; Elsbach, 2006) or may
be viewed as a threat to the
moral authority or legitimacy of the target (Fording, 2001;
Andrews, 2001). Reputation signals
-
11
the quality and reliability of a firms practices, indicates a
firms overall level of prestige in the
market (Rindova, Williamson, Petkova, and Sever, 2005) and
distinguishes a firm from its
competitors (Fombrun and Shanley, 1990; Fombrun, 1996; Whetten
and Mackey, 2002).
Reputation is also linked to the legitimacy of an organization
and its access to other institutional
resources that enhance its competitive ability (Oliver, 1991;
Rao, 1994; Rindova and Fombrun,
1999). When a firms reputation partly rests on its demonstration
of social responsibility,
trustworthiness, and conformity to industry and societal norms
(as most do), the firm is
vulnerable to reputation-based attacks from activists and other
secondary stakeholders (Harrison
and Freeman, 1999). Thus, extrainstitutional tactics may
indirectly influence firms by
threatening to denigrate their public image and constrain access
to institutional resources.
For our purposes (and for managers gauging the potential impact
of a boycott), boycotts may be
most threatening when they are able to communicate their
grievances indirectly through third-
party stakeholders. The media offers one channel that allows
boycotts to activate potential
threat and put pressure on corporate executives (Lipsky, 1968;
Dyck and Zingales, 2002).
Sociologists have argued that social movements use media or
journalistic accounts to challenge
culturally dominant images, such as those offered by
corporations (Gamson, Crouteau, Hoynes,
and Sasson, 1992). Friedman (1999) argues that boycotters
explicitly use media outlets to
publicize their claims and mobilize potential participants. For
example, the Cesar Chavez-led
grape boycotts effectively allied with the media to generate
national excitement about the
boycott, extending the range of influence well beyond the local
migrant worker population and
farm community. Corporate decision-makers are likely to see
sustained media attention to a
boycott as evidence that there is public support for the
boycotters cause and that their reputation
-
12
is being tarnished. Additionally, the media is a potential
conduit of norms of moral and social
appropriateness (Elsbach, 1994; Hoffman and Ocasio, 2001;
Pollock and Rindova, 2003).
Consistent with these institutional explanations of media
influence, boycotters challenge the
legitimacy of a corporate targets practices by indirectly
transmitting grievances through the
media and actively reconstructing the public image of the firm.
Therefore, I expect a positive
association between media attention to a boycott and corporate
concession.
Hypothesis 1: Corporate targets are more likely to concede to
boycotts that
generate high levels of media attention.
Movement opportunities for change
Social movement scholars argue that certain organizations are
more susceptible to being targeted
by activists at a given moment in time. Not all target
organizations are equally vulnerable to
calls for change (e.g. McAdam, 1996; Soule and Olzak, 2004;
Meyer and Minkoff, 2004). Some
corporations may be more open to change than others, and
therefore, more likely to shift their
practices or policies when given sufficient input. In her
examination of corporate policy changes
related to same-sex benefits provision, Raeburn (2004) refers to
these moments of openness to
change as corporate windows of opportunity.
Movement opportunities, in general, may be signals that the time
is ripe to instigate change in a
target organization. One formulation of the hypothesis is that
movement opportunities motivate
activists to target a particular organization. Thus, movement
opportunities are signals to
stakeholders that an organization is ripe for transformation.
However, as Meyer and Minkoff
-
13
(2004) argue, opportunities are not only signals but may also be
structural openings created by
instabilities or changes in elite attitudes that provide a
fertile environment for shifts in
policymaking. In other words, the same factors that make an
organization an attractive target for
movement activists are the same factors that make an
organization susceptible to change.
Therefore, the determinant of success may not necessarily be the
effect of particular activist
tactics; rather, activism merely initiates conversation among
managers who are already
predisposed to make changes.
Although both kinds of opportunities signals of potential
success and structural openings
often overlap, in this paper I focus on the structural openness
of corporations to change. For
concessions to occur, activists must do more than just perceive
a signal of potential
opportunities, the corporate target must actually be
structurally vulnerable. By focusing on
specific structural causes of change, I am also attempting to
more concretely specify the
mechanisms underlying corporate change. Some social movement
scholars have criticized the
political opportunity concept as being vacuous and not
theoretically specific (e.g., Gamson and
Meyer, 1996; Goodwin and Jasper, 1999). Yet, despite these
challenges, social movement
scholars have continued to find explanatory utility in the
concept and have sought to refine it
(e.g., McCammon, Campbell, Granberg, and Mowery, 2001; Soule and
Olzak, 2004; Soule and
King, 2006). In order to provide explanatory leverage, movement
opportunities should be
attached to context-specific mechanisms and should lead to
testable hypotheses about the causes
of change (Meyer and Minkoff, 2004; Cornwall, King, Legerski,
Dahlin, and Schiffman, 2007).
In order to specify the kinds of opportunities that exist for
movements trying to initiate change in
-
14
a corporate setting, I draw on organizational research that
theorizes why some corporations are
more malleable and susceptible to change.
One of the primary sources of change suggested by organizational
scholars is environmental
feedback (Levitt and March, 1988). Negative feedback encourages
managers to reevaluate their
current policies and consider implementing alternatives (Greve,
2003). Managers that receive
negative feedback may feel more freedom to pursue new policies
and practices. One important
source of feedback is performance declines. Specifically, this
means that the firms performance
is floundering relative to expectations that were set by past
performance. Declines in
performance spur change by encouraging managers to remain open
to new input. This line of
thought is compatible with organizational scholars who argue
that instability and uncertainty
within firms spurs political conflict (Jackall, 1988; Morrill,
Zald, and Rao, 2003).
Boycotted firms should be especially likely to respond to
concessions when they experience
declines in the two performance criteria that are (potentially)
directly affected by the threat of
boycotts: sales revenue and reputation. As explained earlier, a
decline in revenue and reputation
represent a potential deterioration of a firms competitive
advantages. Boycotts highlight
problems inherent in companies that are already suffering losses
in sales revenue or reputation.
The boycott becomes the impetus that executives needed to make
course corrections (Garrett,
1987). In both cases, the boycott occurs in an environment where
performance declines generate
managerial willingness to consider alternative means to protect
their tangible and intangible
assets. While the boycott may have been the ultimate cause of
organizational change, the
-
15
proximate cause is the performance decline. Thus, we should
expect that sales and reputation
declines are good predictors of a corporations likelihood to
concede to boycott demands.
Hypothesis 2: Corporate targets are more likely to concede to
boycotts when they
are currently experiencing a decline in sales revenue.
Hypothesis 3: Corporate targets are more likely to concede to
boycotts when they
are currently experiencing a decline in reputation.
The political mediation model of influence
In a series of papers examining legislative changes, Amenta and
his colleagues developed an
approach called the political mediation model (Amenta,
Carruthers, and Zylan 1992; Amenta,
Dunleavy, and Bernstein 1994; Amenta and Young 1999; Amenta,
Caren, and Olasky, 2005;
Amenta, 2006). The political mediation model argues that
movements attempts to influence
policymaking depend on the combination of movement mobilization
and movement
opportunities, or as Amenta states, the productivity of the
collective action of state-oriented
challengers is mediated by political circumstances (2006: 8).
Activists efforts are not equally
effective in all situations. Movement attempts to initiate
change in extremely closed settings
have an attenuated effect on decision-making. But as the context
becomes more malleable and
supportive of change, movement threat is magnified (e.g. Soule
and Olzak, 2004; Soule and
King, 2006). The political mediation model seems particularly
relevant to the study of changes
in corporate policies or practices where changes in response to
secondary stakeholders are
limited due to the relatively closed nature of the corporation.
Thus, assertive, extrainstitutional
tactics, like boycotts, should be most effective when the
conditions are ripe for change.
-
16
Given that the political mediation model was designed to explain
changes in the realm of
governmental politics, where a favorable political context often
means a competitive partisan
system and access to elite allies, it is important to specify
the mechanisms of political mediation
in the corporate context. I focus on two mechanisms. First,
movement threat may accentuate the
immediacy and visibility of the negative feedback. Increasing
threat caused by a boycott raises
awareness of other internal problems and further destabilizes
the power structure that depends on
the legitimacy of their authority for effective governance.
Extensive media coverage of a
boycott brings the declining performance of the company to light
and allows activists and other
stakeholders to reframe the performance in the worst possible
light. The indirect influence of
boycotts intensifies all of the negative perceptions created by
the performance decline.
Performance declines, in this sense, are not interpreted the
same way in every context. This
mechanism also highlights the rhetorical role that movements and
media play in shaping the
perceptions that other stakeholders have of the target
corporation.
Second, the negative feedback created by performance declines
may sensitize managers to the
concerns of stakeholders and, in particular, to the grievances
of boycotters. Research showing
that managers become more active in image management when the
firm has undergone corporate
restructuring or other uncertainty-inducing changes provides
support for the idea that secondary
stakeholders gain more influence when organizations are subject
to negative performance
feedback (Griffin, 2004; Dentchev and Heene, 2004). Managers may
feel vulnerable to outside
criticism when facing an internal crisis and the organization is
exposed to heightened uncertainty
(Jackall, 1988). Executives may feel that performance declines
limit their ability to take a firm
-
17
stance against the activists. Managers may also feel that
performance declines enhance external
monitoring, which sensitizes the executives to their stakeholder
environment.
Lacking confidence in their ability to satisfy various
stakeholder demands, executives of
companies with performance declines become more susceptible to
activists agitation. These
executives fear that additional disruption of stakeholder
sentiment will lead to an even greater
decline in performance. Firms with declining sales may perceive
that a boycott will further
reduce the loyalty of their customer base and cause more revenue
slippage (Garrett, 1987).
Firms experiencing a perceived decline in reputation may view
the threat of boycott as an
additional signal that the company is losing credibility or
legitimacy in the eyes of important
stakeholders. In both cases, firms targeted by boycott see
concession to the boycotters demands
as the best option for correcting (or at least not further
inflaming) the firms sales or reputation
problems. Firms not experiencing sales or reputation declines
will be less likely to see the
boycott as a real threat, and therefore will choose to ignore
the boycott or deal with it in some
other way (e.g. combating the boycotters claims in the
media).
Given these conditions, the effectiveness of movement tactics
will be highly contingent on the
internal malleability of the organization. We should expect a
positive interaction between the
potential threat that the boycott represents (i.e. the amount of
media attention to the boycott) and
corporate opportunities for movement influence (i.e. declines in
sales and reputation).
-
18
Hypothesis 4: Media attention to a boycott will have a stronger
effect on the
likelihood of corporate concessions when the corporate target is
experiencing a
decline in sales.
Hypothesis 5: Media attention to a boycott will have a stronger
effect on the
likelihood of corporate concessions when the corporate target is
experiencing a
decline in reputation.
Data and Methods
Data and dependent variable
To assess the determinants of corporate concessions to boycotts,
I created a dataset of United
States-based boycotts during the 1990-2005 time period that were
levied against publicly traded
corporations and reported in the media. I chose to examine this
group of firms for a couple of
reasons. First, Standard and Poors Compustat database has data
on these firms, which allow me
to assess the impact that firm characteristics (such as decline
in sales) have on the likelihood of a
corporate target concession. Second, this group of firms is also
elite, in the sense that they are
large and successful, and are the kind of organizations that
should be highly sensitive to
shareholder concerns while giving less credence to the demands
made by secondary
stakeholders, like movement activists. If any firm will ignore
social movements claims, it
would likely be publicly-traded corporations for which the
shareholder is normatively the most
important stakeholder group. Thus, looking at the effects of
boycotts on publicly-traded firms in
the United States is a strong test of these hypotheses.
-
19
Data on boycotts came from five national newspapers: the New
York Times, Washington Post,
Wall Street Journal, Chicago Tribune, and Los Angeles Times. The
selection of newspapers was
intended to reduce the potential for regional (especially
Eastern) bias noted in other studies of
boycotts (Friedman, 1985). Research assistants searched for all
articles during the relevant time
period that contained the word boycott in the text of the
article.4 The assistants then read each
article to ascertain that the boycott was targeted against a
corporation, and then the article(s) was
matched to company-specific data. Researchers also coded these
articles to obtain relevant
information about the boycotts. Coding newspaper articles as a
data gathering method is
common for scholars studying corporate boycotts (e.g. Friedman,
1985; Pruitt and Friedman,
1986; Davidson, Worrell, and El-Jelly, 1995). Newspapers are
also a frequent source of data for
scholars studying other social movement tactics, like protest
(e.g. McAdam and Su, 2002; Earl,
Soule, and McCarthy, 2003; Earl, Martin, Soule and McCarthy,
2004; Van Dyke, Soule, and
Taylor, 2004). Using this method I found 133 separate boycotts.
Several boycotts targeted
multiple firms, which makes the total number of corporate
targets of boycotts slightly higher at
189. Due to missing financial data for some of these firms, the
regression analysis below only
reports 144 observations of boycott targets.
A striking feature of the dataset is the broad variety of issues
motivating boycotters. Morality
issues were the most common set of concerns that motivated
boycotts (23%). An example of a
morality issue would be when a group of concerned citizens
boycotted a television network due
to violence or sexuality portrayed in a television series. The
second most common issue was
4 Research assistants searched newspaper article texts using a
combination of several online databases (Proquest,
Factiva, and Lexis-Nexis).
-
20
racial discrimination (13%), followed by environmental concerns
(11%), labor conflict (10%),
and consumer product problems (9%).
The dependent variable in the analysis is binary concession to a
boycotters demands.5 A
concession, according to the criteria of the coding process, is
a recognition by the corporation of
the boycotters demands and a public expression of willingness to
conform to those demands.
The criterion excludes boycott responses that are attempts by
the target to de-mobilize the
activists without making any real changes (e.g. were considering
a change); however, it is
also a possibility that not all corporate concessions achieve
all of the intended goals of the
boycotters. The definition of concession used here is similar to
Friedmans (1985) consequence
criterion for boycott success. According to his definition, a
boycott has positive consequences
if the larger objectives of the boycott were met (1985: 106). To
illustrate, in 1992 a Dallas
police officers organization called for a boycott of Warner
Records for releasing an album by
the rap artist Ice-T that contained a song Cop Killer. Warner
Records eventually relented and
re-released the rap album without the offensive song. This case
represents a concession in that
the corporation specifically gave in to the demands of the
boycotters and made that concession
publicly known.
Information on concessions came from an additional search of
newspaper articles relating to the
boycotts conducted using the Factiva database. Using an
alternate source of information about
the boycotts ensures full coverage of the events. The variable
rests on the assumption that if a
corporate target wants to concede to the boycotters demands,
they will do so publicly, given that
the boycott is a public expression of collective action. A
concession must also be public if it is to
5 Inter-coder reliability rates on all variables in the analysis
were consistently at or above 90% agreement.
-
21
be effective in curtailing the intended effects of the boycott.
Given the timing of a concession is
likely an important indicator of the activists influence, I used
a six-month time window
(beginning with the first announcement of the boycott) to record
all boycott activity and
concessions.6 According to these data, 53 (or 28 percent) of the
corporate targets conceded to
the boycotters demands. This number is similar to the concession
rate of 26 percent found by
Friedmans (1985) study of boycott results.
Independent variables
To test hypothesis 1, I included a variable that captured the
level of media attention granted to a
particular boycott. This variable is operationalized as the
number of newspaper articles
discussing a boycott prior to a concession or occurring during
the six-month time period
following the announcement of the boycott. Articles on boycotts
were found from the
newspaper article search described above. Given that the amount
of media coverage to any
given boycott is partly a function of the overall salience and
visibility of the firm, I include two
control variables in the regression to account for overall firm
visibility firm size (the natural
log of the firms assets) and past media attention given to firm.
Both of these variables are
described in more detail below.
To test hypotheses 2 and 3, I included variables that measure
declines in sales and reputation.
Decline in sales is computed using COMPUSTAT data on net sales.
The variable measures the
change in net sales (in millions) from two years prior to the
boycott to the year in question and is
reverse coded so that a positive value of the variable implies a
larger decline in sales. Thus, a
6 Changing the time window did not alter the percent of
corporate concessions greatly. Most corporations that
concede to boycotters do so within a few weeks of the boycott.
Thus, the six-month time window provides ample
time to assess the possibility of concession.
-
22
positive coefficient for this variable would indicate that a
decline in sales positively influences
the likelihood of concessions. Decline in reputation is computed
using Fortune magazines list
of the Top 100 Most Admired Companies in the United States.
Other organizational scholars
have used the periodicals admiration score as an indicator of a
companys overall reputation
among stakeholders (McGuire, Sundgren and Schneeweis 1988; Staw
and Epstein, 1990;
Fombrun and Shanley 1990; Roberts and Dowling 2002; for a review
of alternative measures of
reputation see Berens and Van Riel, 2004), making this study
consistent with extant research on
reputation. While some have questioned the utility of Fortunes
admiration index, among
reputation measures it is still the most widely used in the
empirical research arena (Sabate and
Puente, 2003: 162). The variable is calculated using an ordinal
transformation of the raw scores,
resulting in four ordinal values. The lowest ordinal value, 0,
signifies that the firm did not
make the top 100 list. The highest ordinal value, 3, indicates
that the firm had a reputation
score of greater than 7.7 The reputation decline variable, then,
is calculated by measuring the
change in reputation value from two years prior to the boycott
to the year in question and, like
the sales decline variable, is reverse coded so that a positive
value implies a larger decline in
reputation. The reputation score is calculated using data from a
survey taken in the prior year,
and so using the score from the year in which the boycott
occurred in effect lags the variable by
one year.
Hypotheses 4 and 5 predict interaction effects between the media
attention variable and the two
variables measuring sales and reputation decline. In order to
reduce collinearity, a common
7 A value of 1 is equal to a raw score ranging from 1 to 5 and a
value of 2 equals a raw score ranging from 5.1 to
7.99. The distribution of reputation scores among boycott
targets is divided among the four categories in the
following way: 0 (38%), 1 (22%), 2 (19%), and 3 (21%).
-
23
problem when including statistical interactions (Jaccard,
Turrisi, Wan 1990), I center the main
variables on their means prior to calculating the interaction
variables.
Control variables
I include a number of control variables in the analysis,
including additional boycott and
corporate target characteristics. Number of social movement
organizations is a count of the
number of activist organizations that are mentioned in the
newspaper articles as being associated
with the boycott. Protest is a dummy variable indicating whether
activists accompanied the
boycott with a public demonstration to air their grievances.
Most research on extrainstitutional
tactics considers protests to be one of the more prominent and
influential tactics of activists (e.g.
Soule et al., 1999; King, Bentele, and Soule, 2007) and may be
used to accentuate the effect of a
boycott. Leafleting is a dummy variable indicating whether
activists distributed fliers or leaflets
to announce the boycott. The presence of leafleting suggests a
higher level of internal
coordination and indicates that the boycott is actually taking
place (compared to boycotts that
merely announce a boycott). Abnormal returns is a dummy variable
indicating whether the
stock price of the target firm exhibited negative abnormal
returns during the time window around
the boycott announcement. We might expect that companies will be
more likely to respond to
boycotts that cause investors to react negatively. The dummy
variable is derived from a measure
of the cumulative abnormal returns to the corporate targets
stock price during an eleven-day
time window (the five days prior to the boycott until the five
days after its announcement)
surrounding the boycott event. Negative abnormal returns
indicate that investors see a particular
event has having potentially negative effects on net present
value of the firm and suggest a loss
of investor confidence. The measure is commonly used in event
studies to assess the effect of
-
24
any exogenous event on a firms stock value (e.g. (Patell, 1976;
Brown and Warner, 1985). The
variable is calculated by taking the sum of the differences of
the observed daily returns from the
expected returns, based on past correlations between the firms
stock price and changes in the
market index.8 I used the Eventus software to create the
variable.
Institutionalized issue is a dummy indicating whether the issue
targeted by boycotters was an
institutionalized policy or practice that would be difficult and
costly to change. In contrast, other
issues, such as those involving product or marketing changes,
might be less institutionalized in
the corporate structure and therefore easier to change in
response to boycotters. I also included
five dummy variables indicating the type or subject matter of
the issue that concerns boycotters.
The five types were environmental, labor, consumer,
discrimination, or morality-related issues.
Presumably, there may be variation in the responsiveness that
corporations exhibit to different
issues.
Multiple targets is a dummy variable that measures whether the
boycott targeted more than one
corporation. Subsidiary is a dummy variable that indicates
whether the target corporation is a
subsidiary of a larger corporation. Presumably managers in
subsidiaries would have less
discretion in deciding how to respond to a boycott than managers
of non-subsidiaries. The
natural log of assets owned by the firm is a measure of
organizational size. Data for this and
other financial variables came from Standard and Poors COMPUSTAT
database. Cash flow is
8 The cumlative abnormal returnjt = Rjt aj bjRmt, where Rjt is
the rate of return for a day around a boycott and aj and bj are
regression coefficients taken from the following expected return
equation: Rjt = j + jRmt + jt, where Rjt
is the rate of return for firm j for a period of days preceding
the boycott announcement, Rmt is the market return (the
equally-weighted daily return for all firms in the CRSP index)
on day t, j is the systematic risk of firm j, j is the
rate of return on firm j when Rmt is zero, and jt is a serially
independent disturbance term with E(jt) = 0. Rjt can be
interpreted as the expected return for the stock of firm j
holding constant shifts in the overall market portfolio. The
regression coefficients for expected return are calculated for a
239-day period prior to the beginning of the event
window.
-
25
a variable that indicates how many excess resources a firm has
that it may use to absorb
disruption costs imposed by boycotts. Cash flow is a firms
operating income plus depreciation
value divided by the firms common shares. Reputation is the
ordinal value of a firms
reputation as determined by Fortunes annual list of the most
admired companies. The
description of how the yearly ordinal values were determined is
described above. Past media
attention measures the number of articles mentioning the
corporate target in the year preceding
the boycott announcement. I control for this to account for the
plausibility that the variable
measuring level of media attention to the boycott may actually
be capturing overall attention
granted to the company. Thus, by including the overall media
effect, we can more accurately
discern the independent effect that media coverage of the
boycott has on likelihood of conceding.
I also include annual time dummies to account for unmeasured
temporal heterogeneity. I do not
show the effects of these time dummies in the regression results
below. Table 1 contains
descriptive statistics and correlations for all of the
independent variables (except those of the
annual time dummies).
[Table 1 about here]
Statistical Analysis
To assess the effects of independent variables on the likelihood
of concession to a boycott, I use
a two-stage Heckman probit model. A probit regression is
appropriate for models where the
dependent variable is a dummy variable. The two-stage model
accounts for sample selection
problems that may bias the results. The Heckman probit model is
a variation of the original
Heckman regression model (Heckman, 1979) in that the
second-stage uses probit regression
-
26
rather than ordinary-least squares regression. However, the
model differs from standard probit
(or logistic regression) due to a first-stage probit regression
that is used to estimate a selection
effect coefficient (known as the inverse Mills ratio or ), which
is then added as a control to the
second-stage model. Substantively, represents the probability
that a corporation will not be
targeted for boycott. Heckman selection models are routinely
used in the management literature
when sample selection bias may be a problem (e.g. Shaver, 1998;
Leiblein, Reuer, and Dalsace,
2002). I used Statas heckprob function to obtain
maximum-likelihood estimates, adjusting the
standard errors by clustering cases at the firm-level.
To conduct the first-stage probit models, I gathered additional
data on companies that activists
did not boycott. A comparable group of firms for which data were
available were the five-
hundred firms with the greatest net sales revenue. This set of
firms is roughly equivalent to the
Fortune 500 list of largest U.S. companies. Given that most of
the firms targeted for boycott
also came from this group of large, publicly-traded companies,
this represented an adequate
group for comparison. I also added two additional independent
variables to the first-stage probit
model: the number of times that a firm has been boycotted in the
past five years and the number
of times that firms in the same industry have been boycotted in
the past five years. For the
second variable the industry is conceived as the firms
four-digit SIC code. These variables
measure the propensity for activists to target a particular firm
or industry for boycott. In this
model, I also included several of the firm-level independent
variables (sales decline, reputation
decline, reputation, and logged assets) to control for other
firm characteristics that might enhance
the firms attractiveness of a target. Although not the
theoretical focus of this paper, the analysis
will indicate factors that predict the likelihood that a
corporation will face boycotts. Thus, the
-
27
model will indicate whether the same factors that predict
openness to change in response to
boycotters are the same factors that predict the likelihood of
being targeted for a boycott. The
underlying assumption of the political opportunity structure
model discussed above is that these
sets of corporate attributes should be the same. The primary
purpose of the analysis, however, is
merely to produce a selection effect coefficient (), which will
facilitate the estimation of
unbiased results in the second-stage regression model.
Results
Table 2 shows the results of the first-stage probit model.
Activists are more likely to boycott
firms that have previously been targets of boycott or that are
in industries that have been targeted
frequently in the past. Activists are also more likely to target
large firms (as demonstrated by the
positive coefficient of logged assets) and firms with positive
reputations.. Interestingly, the
variables predicted to explain openness to change, sales decline
and reputation decline, do not
make corporations more attractive boycott targets to activists.
Rather than focusing on firms that
are in a weaker financial or reputational state, the results
imply that activists tend to boycott
firms with strong reputations and that are highly visible due to
size. Size and corporate
reputation appear to be powerful magnets for attracting unwanted
attention from social
movement activists.
[Table 2 about here]
-
28
Table 3 contains the results of the second-stage probit
regression model in which the dependent
variable is corporate concession to boycotters demands.9 Model 1
provides a test of hypotheses
1-3. The level of media attention to a boycott has a positive
effect on the likelihood of
conceding to boycotters demands, thus providing support for
hypothesis 1. Media attention to a
boycott heightens corporate decision-makers concern that the
boycott is a potential threat. The
model does not provide support for hypothesis 2. Corporate
targets of boycott are no more likely
to concede if they have experienced a sales decline. However,
the model does support
hypothesis 3. Firms that recently experienced reputation
declines are much more likely to
concede to boycotters demands.
[Table 3 about here]
Models 2 and 3 provide tests of hypotheses 4 and 5,
respectively, which are based on the
political mediation model of social movement influence. The
interaction effect of the level of
media attention to the boycott and sales decline is not
statistically significant (model 2).
Therefore, the findings do not support the hypothesis that a
performance decline in sales revenue
moderates the potential threat of the boycott. The interaction
effect of the number of newspaper
articles related to the boycott and reputation decline is
statistically significant. The analysis,
then, supports the hypothesis that reputation decline
accentuates the effect of a boycotts media
attention on the likelihood of corporate concessions.10
9 While the selection effect coefficient () is not statistically
significant, controlling for the effect improves our
confidence that the external and internal validity of the
results are not compromised (Berk, 1983). 10 One might argue that
some reputation levels are more valuable than others and that the
effect of reputation decline
can be explained by either a loss of a once very high reputation
or by transitioning to a very low reputation. To test
this hypothesis I ran regression models including dummy
variables for each kind of transition (replacing the main
reputation decline variable). None of these reputation decline
variables were statistically significant on their own,
supporting the interpretation of the results provided above.
-
29
These findings are presented graphically in Figure 1, expressed
as the expected probability of
conceding. The three lines in the figure track different
probabilities of conceding depending on
how much the firms reputation declined in the past. Firms have a
roughly equal probability of
conceding with only one article of newspaper coverage on the
boycott, but at higher levels of
media coverage the probability changes greatly based on what
kind of reputation decline the firm
previously experienced. Firms one standard deviation above the
mean in reputation decline
rapidly increase in probability at higher levels of media
coverage. Firms at mean levels of
reputation decline experience more linear increases in the
probability of concessions with higher
levels of media coverage. In contrast, firms that experienced
reputation gains have a very low
probability of conceding to boycotters, even at very high levels
of media coverage. These firms
are about as likely to concede when four boycott-related
articles cover the boycott as when there
is only one article in print. Thus, reputation gains buffer
firms from the threat of boycotts. Firms
that declined in reputations, on the other hand, are much more
exposed to boycott threats.11
[Figure 1 about here]
11 I also used a discrete-time event history analysis to assess
the effects of the explanatory variables on the rate of
concessions among targeted organizations. Compared to the probit
regression, the event history analysis determines
the factors that affect the timing of concession, and therefore
it is possible that the timing of the event of concession
is shaped by different causal factors. The event history model
also allows the amount of media attention (and the
media attention interaction variables) to vary over time. The
event history model, however, does not control for the
selection effect of being chosen as a boycott target. The
results of the event history analysis (not shown here but
available upon request) confirm that the amount of media
attention positively affects the timing of concession and
that firms experiencing reputation declines are quicker to
concede to boycotters; however, the interaction effect
between reputation decline and media attention was not
statistically significant. This finding is not entirely
surprising given that a number of other factors likely affect
the timing of concession that have little to do with the
amplification of movement mobilization suggested by the
political mediation effect. Importantly, the results about
concession timing do not negate the main finding that reputation
decline amplifies activist influence on the net
probability of concession, but they do suggest that the timing
of concession is unaffected by this mediating
influence.
-
30
Discussion
Boycotters want to put pressure on firms to change their
practices or policies. A concession to
boycotters demands justifies the use of the tactic. Firms that
give in to boycotts demonstrate
that they are willing to be persuaded by external influences.
Concession to boycotters shows that
firms can be influenced by secondary stakeholders that lack
means of influence through
legitimate channels of organizational change. But not all firms
are equally influenced. As social
movement theorists have forcefully argued, some activist targets
are more susceptible to
influence than others.
The findings of this study support the claim that social
movement and corporate target
characteristics moderate the influence that social movements
have when trying to create
organizational change. Social movement theory suggests that one
of the main sources of activist
influence is in their ability to threaten target institutions
through the use of extrainstitutional
tactics. The results confirm that the level of threat imposed by
a boycott is predictive of
concessions. Boycotters are more likely to exert influence when
the boycott receives a great deal
of media attention. Insofar as media attention indicates the
potential level of threat imposed by
boycotters, we can conclude that the ability of boycotts to
threaten the tangible and intangible
resources of the target organization is one reason they are
effective at initiating organizational
change. The findings also suggest that the effectiveness of a
movements tactic is also
moderated by the past performance of the corporate target.
Corporations experiencing a
reputation decline are more likely to take the threat of a
boycott seriously. On the other hand,
corporations that experienced gains in reputation feel protected
from attacks made by activists
-
31
and are therefore more likely to stick to their guns even if the
apparent threat level of the
boycott is high.
The study also speaks to potential mechanisms that facilitate
boycott threat. Whereas boycotts
are often thought to be effective due to their ability to
constrain sales revenue (Friedman, 1999;
Vogel, 2005), these findings suggest that the most critical
mechanism underlying boycotts
influence is their ability to damage corporate reputations. As a
threat to reputation, boycotts
challenge the legitimacy of a firms practices and damage the
overall prestige of the target
organization. The analysis indicates that previous sales decline
had no effect on the likelihood of
concession, nor did sales decline moderate the effect of media
attention on the likelihood of
concession. In comparison, reputation decline both directly
affected a targets propensity to
concede and caused corporations to be more sensitive to the
potential threat of the boycott.
Therefore, it is reasonable to assume that corporate
decision-makers viewed boycotts as a more
serious threat to their reputation than to their sales revenue.
The finding helps make sense of
previous studies that have questioned why boycotts are ever
effective given that most boycotts
do not involve a large number of participants and do not usually
have a large impact on the sales
of the corporate target (Vogel, 2005). Boycotts may not need to
affect sales at all in order to be
effective. Rather, boycotters influence stems from their ability
to make negative claims about
the corporation that generate negative public perceptions of the
corporation. Hence, corporations
that are already struggling to maintain their previously
positive reputations will be more likely to
concede to boycotts and quell any further damage the boycott may
do to their reputation.
Conclusion
-
32
This study offers potential contributions to both the study of
social movement outcomes and
organizational theory. In this section I outline several
theoretical insights provided by the study
and suggest avenues for future research in both literatures.
Implications for the study of movement outcomes
While past research focused on the overall movement effort to
exert influence in the corporate
realm (Zald and Berger, 1978; Hoffman, 1999; Lounsbury, 2001;
Schneiberg, 2002; Scully and
Segal, 2002; Raeburn, 2004), the contribution of this study is
to focus on the determinants of the
effectiveness of one particular kind of extrainstitutional
tactic the boycott. Few social
movement studies of any kind, in fact, examine the consequences
of a specific tactic (although,
see McAdam, 1983). Although much social movement research has
discussed boycotts as a
primary tactic, very little attention has been paid to the
determinants of boycott outcomes
(although see, Jenkins and Perrow, 1977; Luders, 2006). More
generally, the study indicates
when extrainstitutional tactics are likely to succeed in
initiating change and how these tactics
generate influence. Rather than disrupting local routines, as
extrainsitutional tactics are often
conceived (e.g., Rojas, 2006; Luders, 2006), corporate boycott
influence is felt indirectly through
the media. Because of the medias position as a conduit of
information and norms, social
movements rely on the media to communicate their grievances and
to actively reconstruct the
publics perceptions regarding the appropriateness of a firms
behavior. In this way, the media
amplifies the threat of a boycott.
The study also provides support for the political mediation
model (Amenta, Carruthers, and
Zylan 1992; Amenta, Dunleavy, and Bernstein 1994; Amenta and
Young 1999; Cress and Snow
-
33
2000; Amenta, Caren, and Olasky, 2005) and extends it by
demonstrating that the contextualized
influence of activists is also apparent in the realm of
corporate politics. Following Amentas call
to specify what constitutes a favorable political context and
what does not (2006: 6), this
formulation of the political mediation model identifies the
conditions in a corporate setting
where managers are most likely to be receptive to movements
attempts to initiate change. By
focusing on the context of legislative politics, political
mediation theorists have theorized
institutional conditions in a relevant but also narrow way.
Broadening the discussion of
institutional change to encompass movement targets that have
alternative change mechanisms
forces us to consider other ways in which movements may
influence elite decision-makers.
Movements targeting corporate actors, for example, may rarely
negotiate directly with or lobby
corporate executives to achieve change (as they often do with
legislators); rather the force of
their indirect influence through challenges to the corporate
image in the media may be sufficient
to instigate change processes. Media attention to a boycott
shapes public perceptions of a target
firm. As media coverage draws increased and unwanted attention
to a corporation, the internal
conflicts within a target may be amplified, thereby loosening
the grip of incumbent elites.
More generally, these findings challenge us to specify the
mechanisms whereby activists attain
political influence in any setting where elite allies are not
present. Activists are most likely to
resort to the use of extrainstitutional tactics, like boycotts,
when trying to gain influence in
relatively closed context with few allies.12 Therefore, what
opportunities for influence exist
12 Amenta (2006; Amenta, Caren, and Olasky, 2006) also makes
this point by contending that assertive strategies, of
which extrainsitutional tactics is a subset, are most needed in
closed contexts. However, one of the implications of
the political mediation model is that other sorts of strategies
may be better suited when targeting corporations in
different environments or providing different opportunities for
activists. For example, if a movement already has
allies on a corporate board, boycotting may be the least
effective tactic for generating change and direct talks with
executives may be more appropriate. To extend the political
mediation model further, future research should
-
34
when few or no elite allies step forward to support the cause of
activists? Other studies have
shown that, even if direct elite ties are not present, the
restructuring of power relations or
competition between different elite factions may facilitate
movement influence (Soule and
Olzak, 2004; Soule and King, 2006; Olzak and Ryo, 2007). This
study builds on that idea in that
disruptions among powerful factions in an organization may be
one reason that negative
feedback elicits greater attention and sensitivity to secondary
stakeholders (Jackall, 1988). When
needed resources, such as reputation, are suddenly diminished,
elite decision-makers become
exposed to their political and social environment and are more
vulnerable to external attacks.
Thus, resource instability is a key opportunity for generating
movement influence. Future
research of social movement outcomes should consider how this
causal pathway might also
operate in non-corporate settings, given that corporations are
not the only organizations
dependent on resources for functioning and survival.
The analysis also demonstrates that the factors that motivate
activists to target particular
corporations for boycotts are not always the same things that
make a corporation open to change
and susceptible to activist influence. Activists tend to target
large firms with positive
reputations. Neither of those corporate characteristics,
however, predict the likelihood of boycott
success. Instead, it appears that the firms that already have
received negative feedback from
their stakeholder environment, via reputational decline, are the
most susceptible. Thus, activists
interpretations of signals in the corporate environment are not
consistent with the structural
opportunities for change. This finding speaks to social movement
theorists who argue that
political opportunities as signals do not always match the
structural political opportunities that
examine the multiple pathways that activists targeting
corporations might take to generate influence. This more
extensive treatment of the political mediation model, however,
falls outside the scope of this paper.
-
35
facilitate influence (Meyer and Minkoff, 2004; Cornwall et al.,
2007). This finding may also
explain why activists seem to have a fairly low rate of success.
If activists tend not to target the
most structurally susceptible firms, they may often set
themselves up for failure.
Implications for organizational theory
One of the problems facing organizational scholarship is to
generate an explanation for the
activation of influence of secondary stakeholders, who may have
a different view of the
organization and make different claims on resource distribution
than managers (Clarkson, 1995;
Frooman, 1999). The study supports, more broadly, a political
economy perspective of the
firm in which external sources of power are conceived as capable
of shaping organizational
decision-making and instigating change (Zald, 1970). Attempts to
theorize the firm from a
stakeholder perspective (e.g. Donaldson and Preston, 1995) and
the recent trend to import social
movement theories of mobilization and influence reflect a need
to better conceptualize how sets
of external actors attempt to exercise influence within the
corporate realm (e.g. Lounsbury, 2001;
Davis et al., 2005). This study provides insight into one of the
mechanisms for possible
influence when more legitimate channels of change have been
closed to stakeholders. Social
movement tactics, like boycotts, are an alternative means of
influence. These tactics expand the
range of activities that corporate actors must pay attention to
when setting goals, policies, or
building new practices. Inasmuch as social movements have the
ear of the public, they are
capable of transforming the very institutional environment in
which they and corporations
operate. In this sense, social movements take the role of
extra-institutional entrepreneurs
(King and Soule, 2007).
-
36
In a real way, stakeholders are capable of imposing certain
limits on corporate autonomy. Legal
and corporate governance scholars often argue that shareholder
well-being has primacy in the
decision-making of corporate elites (e.g. Smith, 1998). While
not negating this basic
proposition, the findings complicate the question by introducing
alternative contingencies that
executives deal with when deciding how to best manage investor
expectations. As stakeholder
theorists have argued, firms must deal with a variety of
stakeholder groups in their effort to
secure valuable resources. Efforts to satisfy various
stakeholders are sometimes manifest in
concerns about corporate reputation and legitimacy. Because
reputation is rare and inimitable
(Barney, 1991; 2001), is a source of differentiation and
competitive advantage (Fombrun and
Shanley, 1990) and has been shown to enhance organizations
financial value (Fombrun, 1996;
Roberts and Dowling, 2002), firms do not want to tarnish their
reputation and lose this intangible
resource. Inasmuch as reputations are related to institutional
legitimacy and claims to moral
appropriateness (Rao, 1994), firms may also be sensitive to
stakeholder attacks that they perceive
as threatening their access to resources that are contingent
upon institutional support (DiMaggio
and Powell, 1983; Oliver, 1991; Suchman, 1995). Firms are
conscious of the need to manage the
perceptions of important stakeholder groups in the
interpretational domain of their strategic
endeavors (Rindova and Fombrun, 1999: 697). Thus, corporate
concessions to boycotters should
not be interpreted as an irrational reaction to a firms
environment; rather, corporate decision-
makers may be well aware of possible negative financial
consequences resulting from further
erosion of their reputations. As reputation declines, any
financial value based in this intangible
resource may also deteriorate.
-
37
The implication of this logic is that some firms are much more
dependent on secondary
stakeholders due to their risk of losing these critical
institutional resources (Oliver, 1997;
Frooman, 1999). A firms autonomy is limited by its dependence on
certain stakeholder groups
to sustain certain resource bases (Pfeffer and Salancik, 1978).
Firms that benefit from their
corporate reputation may, in a certain sense, be held hostage by
this source of success. While
some have suggested that reputation has a halo effect for
organizations, particularly during times
of crisis or poor financial performance (Bromley, 1993; Fombrun,
1996; Coombs and Holladay,
2006), this analysis suggests that reputation may sometimes
weaken an organizations ability to
deal with external events and crises. Firms become more
vulnerable to activists when they have
positive reputations. They must maintain those reputations in
order to not lose their autonomy
over decision-making to their stakeholders. Firms that suffer
declines in their formerly positive
reputations become increasingly dependent on the satisfaction of
secondary stakeholder groups.
Extrapolating further, organizations that lose control over
intangible assets, such as reputation or
legitimacy, become increasingly constrained by their
institutional environment, of which
stakeholders are one component. Thus, this study provides
evidence for how actors sometimes
perceived to be marginal in an organizational field shape and
constrain their more powerful
counterparts.
-
Table 1: Descriptive statistics and correlation matrix Variables
Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1 Concession .31 .47 1
2 Media attention (mean-centered) .02 1.67 .28 1
3 Sales decline (mean-centered) .00 5623 .07 -.02 1
4 Reputation decline (mean-centered) .18 .56 .12 -.04 .08 1
5 Sales * media -.07 2.09 -.17 -.67 -.21 -.05 1
6 Reputation * media -.06 1.42 .09 -.003 -.06 .12 -.18 1
7 # of SMOs 3.13 6.67 -.05 -.08 .12 -.15 -.003 .09 1
8 Protest .002 .05 -.03 .02 .02 .01 .04 .09 -.13 1
9 Leafleting .00 .03 -.07 .18 -.09 -.05 .07 -.29 -.07 .24 1
10 Abnormal returns .99 .10 .12 .07 .01 .06 -.15 .13 .06 .01
-.14 1
11 Institutionalized feature .01 .08 -.01 -.03 -.17 .09 -.03 .07
-.01 .06 .16 -.06 1
12 Environmental .002 .05 -.08 -.10 -.17 .19 .11 -.04 -.06 .01
-.07 .08 -.04 1
13 Labor .002 .04 -.07 .01 .01 .08 .01 .05 -.10 .23 .40 .01 .28
-.11 1
14 Consumer .00 .03 .12 .003 .02 -.10 .02 .04 .06 -.07 -.04 .04
.04 .06 -.06 1
15 Discrimination .003 .05 .09 .05 -.03 .02 -.20 .09 -.04 .04
.02 .03 .44 -.13 .01 -.07 1
16 Morality .004 .06 .05 .09 .04 -.09 -.08 .08 .08 -.13 -.10
-.02 -.26 -.17 -.16 -.09 -.19
17 Multiple targets .007 .09 -.33 -.12 .14 -.03 .03 .06 .11 -.16
.06 .03 -.02 .12 -.06 -.15 -.002
18 Subsidiary .38 .49 .21 .02 .02 -.05 -.06 -.15 -.03 -.21 -.09
-.03 -.04 -.08 -.02 .01 .05
19 Logged assets 8.34 1.07 -.01 .10 -.34 .07 .01 .02 -.10 .06
.06 -.03 .02 .16 .02 .01 -.01
20 Cash flow 8.44 87.8 .12 -.03 -.02 .03 .02 -.01 -.03 -.03 -.02
-.09 -.06 -.03 -.03 -.02 -.03
21 Reputation 1.41 1.17 -.02 .19 -.23 -.37 -.10 -.02 .11 .01 .10
-.06 .002 -.03 .10 -.003 -.07
22 Past media attention 50.3 70.9 .05 .19 -.06 -.20 .21 .10 .07
-.06 -.09 .02 .03 -.07 -.07 -.03 .04
23 Past boycotts of firm .06 .34 .15 .23 -.07 .10 -.11 .06 .12
-.01 -.10 -.01 -.01 -.07 -.12 -.06 -.04
24 Past boycotts in industry .12 .62 .20 .27 -.08 .10 -.12 .05
.08 .01 -.09 -.05 .003 -.01 -.04 -.09 -.06
-
Table 1, continued Variables 16 17 18 19 20 21 22 23 24
16 Morality 1
17 Multiple targets -.20 1
18 Subsidiary .07 .02 1
19 Logged assets -.15 -.08 -.02 1
20 Cash flow .17 -.07 -.05 .11 1
21 Reputation .05 -.10 -.23 .32 .11 1
22 Past media attention .25 .02 .02 .23 -.05 .29 1
23 Past boycotts of firm .15 -.11 .11 .18 -.04 .26 .27 1
24 Past boycotts in industry .06 -.13 .12 .18 -.05 .25 .19 .87
7
-
40
Table 2: Probit estimates for first-stage boycott target
model
Variables Coefficients Robust standard errors
Constant
-4.04*** .80
Past boycotts of firm .54*** .07
Past boycotts in industry .18*** .03
Sales decline
.02 .05
Reputation decline
-.06 .06
Logged assets
.24* .10
Reputation
.25*** .06
Observations 6672
Log pseudolikelihood -555.90 ***p
-
41
Table 3: Probit estimates of influences on the likelihood of
corporate concessions to
boycotts, 1990-2005
Variables Coefficients/(Robust standard errors)
Model 1
Model 2 Model 3
Constant -.19
(1.25)
-.33
(1.35)
-.19
(1.85)
Media attention to boycott .62*
(.24)
.60**
(.22)
1.87***
(.39)
Sales decline .07
(.22)
-.009
(.17)
.22
(.24)
Reputation decline .63**
(.22)
.61**
(.57)
.73**
(.25)
Articles * Sales decline
-.23
(.32)
Articles * Reputation
Decline
.88***
(.15)
Control variables
# of social movement
organizations
-.002
(.004)
.0002
(.004)
-.003
(.004)
Protest
-.46
(.38)
-.52
(.40)
-.87
(.49)
Leafleting
-1.37
(1.56)
-1.53
(1.65)
-.88
(.97)
Abnormal returns
.28
(.24)
.27
(.34)
.47
(.39)
Institutionalized feature .005
(.40)
-.01
(.41)
.38
(.46)
Environmental issue -.57
(.59)
-.61
(.57)
-.26
(.64)
Labor issue -.04
(.69)
-.04
(.69)
.78
(.81)
Consumer issue .31
(.79)
.32
(.80)
-.79
(.86)
Discrimination issue .46
(.46)
.52
(.51)
-.18
(.54)
Morality issue -.19
(.49)
-.21
(.49)
-.66
(.53)
Multiple targets
-2.19***
(.44)
-2.31***
(.56)
-3.85***
(.74)
-
42
Subsidiary
1.17**
(.37)
1.19**
(.39)
2.22***
(.51)
Logged assets
-.10
(.09)
-.09
(.10)
-.12
(.12)
Cash flow
.06
(.03)
.06
(.04)
.11
(.04)
Reputation
.12
(.16)
.11
(.16)
.14
(.23)
Past media attention .0003
(.0004)
.0004
(.0004)
.0003
(.0005)
Selection correction effect
()
.10
(.25)
.12
(.25)
.18
(.37)
Observations 143 143
143
Log pseudolikelihood -555.89 -555.73 -544.80
***p
-
43
Figure 1. Predicted probability of concession to boycott based
on probit regression results
0
.2
.4
.6
.8
1
Pr(Concession)
1 2 3 4
Number of articles
Reputation decline at mean
Reputation decline at +1 SD
Reputation decline at -1 SD
-
44
References
Amenta, E. 2006.
When movements matter: The townsend plan and the rise of social
security. Princeton, NJ:
Princeton University Press.
Amenta, E., Caren, N., & Olasky, S. J.
2005. Age for leisure? political mediation and the impact of the
pension movement on U.S. old-
age policy. American Sociological Review, 70: 516-538.
Amenta, E., Carruthers, B. G., & Zylan, Y.
1992. A hero for the aged? the towsend movement, the political
mediation model, and U.S. old-
age policy, 1934-1950. American Journal of Sociology, 98:
308-339.
Amenta, E., Dunleavy, K., & Bernstein, M.
1994. Stolen thunder? huey long's share our wealth, political
mediation, and the second new
deal. American Sociological Review, 59: 678-702.
Amenta, E., & Young, M. P.
1999. Making an impact: Conceptual and methodological
implications of the collective goods
criterion. In M. Giugni, D. McAdam, & C. Tilly (Ed.), How
movements matter: 22-41.
Minneapolis, MN: University of Minnesota Press.
Andrews, K.
2001. Social movements and policy implementation: The
mississippi civil rights movement and
the war on poverty, 1965-1971. American Sociological Review, 66:
71-95.
Barney, J.
1991. Firm resources and sustained competitive advantage.
Journal of Management, 17: 99-120.
Baron, D. P.
2003. Private politics. Journal of Economics and Management
Strategy, 12: 31-66.
Basdeo, D. K., Smith, K. G., Grimm, C. M., Rindova, V. P., &
Derfus, P. J.
2006. The impact of market actions on firm reputation. Strategic
Management Journal, 27: 1205-
1219.
Berens, G., & Van Riel, Cees B.M.
2004. Corporate associations in the academic literature: Three
main streams of thought in the
reputation measurement literature. Corporate Reputation Review,
7(2): 161-178.
Berk, R. A.
1983. An introduction to sample selection bias in sociological
data. American Sociological
Review, 48(3): 386-398.
Berle, A. A., & Means, G. C.
-
45
1932. The modern corporation and private property. New York:
Harcourt, Brace, and World.
Berry, G. R.
2003. Organizing against multinational corporate power in cancer
alley: The activist community
as primary stakeholder. Organization and Environment, 16(1):
3-33.
Bromley, D. B.
1993. Reputation, image, and impression management. Chichester,
UK: Wiley.
Brown, S. J., & Warner, J. B.
1985. Using daily stock returns: The case of event studies.
Journal of Financial Economics, 14:
3-31.
Burstein, P., & Linton, A.
2002. The impact of political parties, interest groups, and
social movement organizations on
public policy: Some recent evidence and theoretical concerns.
Social Forces, 81(2): 380-408.
Clarkson, M. B. E.
1995. A stakeholder framework for analyzing and evaluating
corporate social performance.
Academy of Management Review, 20(1): 92-117.
Coleman, C., Nee, L. D., & Rubinowitz, L. S.
2005. Social movements and social-change litigation: Synergy in
the montgomery bus protest.
Law and Social Inquiry, 30(4): 663-736.
Conard, A.
1988. Beyond managerialism: Investor capitalism? University of
Michigan Journal of Law
Reform, 22: 117-178.
Coombs, W. T., & Holladay, S. J.
2006. Unpacking the halo effect: Reputation and crisis
management. Journal of Communication
Management, 10: 123-138.
Cornwall, M., King, B. G., Legerski, E. M., Dahlin, E. C., &
Schiffman, K. S.
2007. Signals or mixed signals: Why opportunities for
mobilization are not opportunities for
policy reform. Mobilization, 12(3): 195-210.
Cress, D., & Snow, D. A.
2000. The outcomes of homeless mobilization: The influence of
organization, disruption,
political mediation, and framing. American Journal of Sociology,
105(4): 1063-1104.
Davidson, W. N., Iii, Worrell, D. L., & El-jelly, A.
1995. Influencing managers to change unpopular corporate
behavior through boycotts and
divestitures: A stock market test. Business & Society,
34(2): 171-196.
Davis, G. F., McAdam, D., Scott, W. R., & Zald, M. N.
2005. Social movements and organizational theory. Cambridge:
Cambridge University Press.
-
46
Deephouse, D. L.
2000. Media reputation as a strategic resource: An integration
of mass communication and
resource-based theories. Journal of Management, 26(6):
1091-1112.
Deephouse, D. L., & Carter, S. M.
2005. An examination of differences between organizational
legitimacy and organizational
reputation. Journal of Management Studies, 42(2): 329-360.
Dentchev, N. A., & Heene, A.
2004. Managing the reputation of restructuring corporations:
Send the right signal to the right
stakeholder. Journal of Public Affairs, 4(1): 56-72.
Donaldson, T., & Preston, L. E.
1995. The stakeholder theory of the corporation: Concepts,
evidence, and implications. Academy
of Management Review, 20(1): 65-91.
Dyck, A., & Zingales, L.
2002. The corporate governance role of the media. NBER Working
Paper 9309, Cambridge MA.
Earl, J., Martin, A., Soule, S. A., & McCarthy, J.
2004. The use of newspaper data in the study of collective
action. Annual Review of Sociology,
30: 65-80.
Earl, J., Soule, S. A., & McCarthy, J.
2003. Protest under fire? explaining the policing of protest.
American Sociological Review, 68:
581-606.
Elsbach, K. D.
1994. Managing organizational legitimacy in the california
cattle industry: The construction and
effectiveness of verbal accounts. Administrative Science
Quarterly, 39: 57-886.
Elsbach, K. D., & Kramer, R. M.
1996. Members' responses to organizational identity threats:
Encountering and countering the
business week rankings. Administrative Science Quarterly, 41(3):
442-476.
Fligstein, N.
2001. The architecture of markets . Princeton, NJ: Princeton
University Press.
Fombrun, C. J., & Shanley, M.
1990. What's in a name? reputation building and corporate
strategy. Academy of Management
Journal, 33: 233-258.
Fombrun, C. J.
1996. Reputation : Realizing value from the corporate image.
Boston, Mass.: Harvard Business
School Press.
-
47
Fording, R. J.
1997. The conditional effect of violence as a political tactic:
Mass insurgency, electoral context,
and welfare generosity in the american states. American Journal
of Political Science, 41: 1-29.
Friedman, M.
1962. Capitalism and freedom. Chicago, IL: University of Chicago
Press.
Friedman, M.
1985. Consumer boycotts in the united states, 1970-1980:
Contemporary events in historical
perspective. Journal of Consumer Affairs, 19(1): 96-117.
Friedman, M.
1999. Consumer boycotts: Effecting change through the
marketplace and the media. New York:
Routledge.
Frooman, J.
1999. Stakeholder influence strategies. Academy of Management
Review, 24(2): 191-205.
Gamson, W.
1990. The strategy of social protest (2nd ed.). Belmont, CA:
Wadsworth.
Gamson, W., Croteau, D., Hoynes, W., & Sasson, T.
1992. Media images and the social construction of reality. , 18:
373-393.
Gamson, W., & Meyer, D. S.
1996. Framing political opportunity. In D. McAdam, J. D. M