4 a Political Mediation Model of Corporate Response to Social Movement Activism

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