Content uploaded by Jessica A. Kennedy
Author content
All content in this area was uploaded by Jessica A. Kennedy on Feb 21, 2017
Content may be subject to copyright.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
1
Hierarchical Rank and Principled Dissent:
How Holding Higher Rank Suppresses Objection to Unethical Practices
Jessica A. Kennedya**
aOwen Graduate School of Management
Vanderbilt University
401 21st Avenue South
Nashville, TN 37203
E-mail: jessica.kennedy@owen.vanderbilt.edu
Cameron Andersonb
bHaas School of Business
University of California, Berkeley
545 Student Services Building #1900
Berkeley, CA 94720
Email: anderson@haas.berkeley.edu
** Indicates the corresponding author. Please direct all correspondence related to this article to
Jessica Kennedy at the email address listed above.
Author Note
We would like to thank Barry Staw, Jennifer Chatman, Rob Willer, Ming Leung, Toby
Stuart, Nathan Pettit, and Gavin Kilduff for their insightful comments regarding earlier versions
of this manuscript. This research was supported in part by research grants from the Haas
Behavioral Lab at the University of California, Berkeley, the Wharton Behavioral Lab, and the
Zicklin Center for Business Ethics at the University of Pennsylvania. We are grateful for this
financial support as well as for the research assistance of Caitlin Cuan, Sharon Holmes, Anjali
2
Joy, Dane Johnston, Morgan Lemonidis, Maxine Rodenhuis, Caitlin Schultz, Betsy Johnson, and
Robin Eller.
Hierarchical Rank and Principled Dissent:
How Holding Higher Rank Suppresses Objection to Unethical Practices
Abstract
When unethical practices occur in an organization, high-ranking individuals at the top of the
hierarchy are expected to stop wrongdoing and redirect the organization to a more honorable
path—this is, to engage in principled dissent. However, in three studies, we find that holding
high-ranking positions makes people less likely to engage in principled dissent. Specifically, we
find that high-ranking individuals identify more strongly with their organization or group, and
therefore see its unethical practices as more ethical than do low-ranking individuals. High-
ranking individuals thus engage less in principled dissent because they fail to see unethical
practices as being wrong in the first place. Study 1 observed the relation between high-rank and
principled dissent in an archival data set involving more than 11,000 employees. Studies 2 and 3
used experimental designs to establish the causal effect of rank and to show that identification is
the key mechanism underlying it.
Keywords: Power; Status; Hierarchy; Ethics; Principled Dissent; Identification
HIERARCHICAL RANK AND PRINCIPLED DISSENT
3
It is difficult to maintain ethical behavior in an organization. Many factors common to
organizations—including the prevalence of goals, the use of groups to make decisions, and the
common emphasis on money—can encourage unethical behavior (Cohen, Gunia, Kim-Jun, &
Murnighan, 2009; Kouchaki, Smith-Crowe, Brief, & Sousa, 2013; Moore & Gino, 2013; Pillutla
& Chen, 1999; Schweitzer, Ordonez, & Douma, 2004). Often, unethical behavior proceeds
without interruption because people do not perceive ethical problems (Tenbrunsel & Smith-
Crowe, 2008; Warren & Smith-Crowe, 2008). For instance, people can be coopted into unethical
behavior when it is embedded in organizational routines (Ashforth & Anand, 2003) or ordered by
an authority (Milgram, 1963).
High-ranking individuals at the top of organizational hierarchies play an important role in
stopping unethical behavior (Mayer, Nurmohamed, Trevino, Shapiro, & Schminke, 2013). The
ethical standards they set trickle down to affect others’ behavior (Mayer, Kuenzi, Greenbaum,
Bardes, & Salvador, 2009). In fact, people across national cultures view those at the top of
organizational hierarchies as responsible for ensuring ethical decision-making among those they
oversee (Baumhart, 1961; Brenner & Molander, 1977; Hamilton & Sanders, 1995; Sims, 1992;
Sims & Brinkman, 2002).
It is unclear from existing research whether occupying high-ranking positions enables or
disables fulfillment of these responsibilities, however. High-ranking positions typically confer
both power (i.e., control over resources) and status (i.e., respect and deference from others)
(Fragale, Sumanth, Tiedens, & Northcraft, 2012; Tost, 2015; Weber, 1948). The power that
accompanies high-ranking positions enables individuals to form confident moral judgments
despite the fact that many ethical decisions are ambiguous (Flynn & Wiltermuth, 2010;
Wiltermuth & Flynn, 2013). Additionally, powerful individuals are less easily influenced by
HIERARCHICAL RANK AND PRINCIPLED DISSENT
4
others and by social situations (Galinsky et al., 2008; See, Morrison, Rothman, & Soll, 2011).
Therefore, when unethical social influences emerge in organizations, they would seem less likely
to lead powerful individuals astray to the same extent as others (Pitesa & Thau, 2013). Similarly,
the status that accompanies high-ranking positions often leads to morally upstanding behavior,
such as greater fairness (Blader & Chen, 2012). Objecting to unethical practices could be another
type of morally upstanding behavior elevated by holding higher rank.
Why, then, do high-ranking individuals at the top of organizational hierarchies so often
fail to stop unethical practices in their organizations? History is full of such cases. From
accounting fraud (McCullagh, 2006; Patsuris, 2002) and the sale of harmful products (Motavalli,
2010), to health-hazardous overwork (Michel, 2011) and pervasive incivility and discrimination
(Antilla, 2002), many unethical practices persist unchecked in organizations, even when those in
high-ranking positions are likely aware of them.
In the current paper, we attempt to solve this puzzle by studying how occupying a high-
ranking position in an organization’s hierarchy affects the likelihood of engaging in principled
dissent, which is an individual’s effort to protest or change ethically objectionable practices
(Graham, 1986, p. 2). We propose that occupying a position of high rank, ironically, can make
individuals less likely to oppose unethical practices within the organization, as compared to
individuals who occupy positions of lower rank. We propose that this can occur in part because
those at the top of the hierarchy identify with the organization more strongly, and identification
leads them to view the organization’s practices as more ethical. Consequently, although high-
ranking individuals are better enabled psychologically and politically to engage in principled
dissent, they may fail to do so because they see no problem with their organization’s unethical
practices in the first place.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
5
To test these hypotheses, we conducted three studies. Based on previous
recommendations (Chatman & Flynn, 2005), we wanted to test the key relation between
hierarchical rank and principled dissent in both field and laboratory settings. However, studying
principled dissent in the field introduces numerous challenges. It requires an adequately sized
sample of people who have observed unethical behavior, a context in which people will talk
openly about their response to the unethical activity that transpired, and because we are
interested in high-ranking individuals at the top of organizational hierarchies, a clear index of
hierarchical position. Fortunately, we obtained access to an archival data set that met these
conditions. Study 1 examines the relation between hierarchical position and principled dissent in
a field setting of over 11,000 U.S. federal government agency employees. Study 2 then uses an
experimental design to replicate the finding from Study 1, establish the causal role of
hierarchical rank, and test identification as one possible psychological mechanism. In Study 3,
we manipulated the conditions that would enable (or impede) the effects of rank on
identification. Specifically, we created conditions under which participants were unlikely to
identify with their group. In these conditions, we expected rank not to affect principled dissent.
Additionally, the third study tests whether higher rank changes ethical views, preventing high-
ranking individuals from detecting ethical problems. Together, our studies describe why and
when holding a high-ranking position leads to less principled dissent. Our designs thus
complement one another, in that Study 1 tests whether the key relation emerges in the real world,
and later studies ensure that hierarchical rank is the causal variable and establish identification as
one important intervening psychological processes.
Graham (1986) noted that principled dissent can take a variety of forms, including
constructive criticism, protest expressed to others within the organization, reports to audiences
HIERARCHICAL RANK AND PRINCIPLED DISSENT
6
outside the organization, blocking actions, and resignation accompanied by an explanation. Our
studies thus operationalize principled dissent in two ways. Study 1 examines the reporting of
unethical practices. Studies 2, 3, and 4 assess whether individuals express disagreement with
their group’s unethical decision. Both dependent measures represent an effort to change a
morally objectionable pattern of behavior going on in a group and thus are forms of principled
dissent.
Our research makes at least three important theoretical contributions. First, we examine
how hierarchical rank affects principled dissent. By doing so, we test a seminal idea from
Graham’s (1986) influential and widely cited theoretical statement on principled dissent. In
contrast to our hypothesis, she predicted that holding a higher level in the organizational
hierarchy would increase, not decrease, principled dissent. We elaborate on her position below.
To date, this idea has not been directly tested. Second, we provide a rare empirical test of the idea
that identification promotes acceptance of unethical behavior, as proposed in prior theoretical
work (Dukerich, Kramer, & Parks, 1998). Like other scholars (e.g., Smith-Crowe & Warren,
2014), we focus on situations where unethical practices already exist in an organization. We are
interested in whether high-ranking people oppose versus accept unethical practices. Third, we
examine the effects of hierarchical rank on identification with the group or organization – a
relationship that has received little empirical attention (for an exception, see Willer, 2009).
Hierarchical Rank
Our primary aim in this research was to understand why individuals at the top of
organizational hierarchies fail to stop unethical practices as often as they do. Accordingly, our
focus throughout the paper is on rank in an organizational or group hierarchy. Higher rank is
structural (Tost, 2015) and it typically involves greater power, or control over resources
HIERARCHICAL RANK AND PRINCIPLED DISSENT
7
(Emerson, 1962), as well as higher social status, or respect and admiration (Weber, 1948). For
instance, higher-ranking individuals typically have more control over valued resources such as
budgets, the ability to hire and promote others, and discretion over key decisions. Additionally,
higher-ranking people are prominent and typically well-respected. Others seek them out for
advice, admire them, and attend closely to their opinions, for example. Consequently, high-
ranking individuals typically wield a great deal of influence (Tost, Gino, & Larrick, 2013).
Scholars have rightly pointed out that different components of hierarchical rank, such as
power and status, are separate constructs that can be distinguished conceptually and empirically
(Magee & Galinsky, 2008). For example, sometimes people with a high level of power might not
have high status in the eyes of others (cf. Anicich, Fast, Halevy, & Galinsky, 2015; Fast, Halevy,
& Galinsky, 2012), and sometimes people with a high level of status might not possess a great
deal of power (cf. Fragale, Overbeck, & Neale, 2011). However, hierarchical rank in most
organizations and groups includes both power and status (Tost, 2015), and the two variables
correlate with each other very highly (Bales, Strodtbeck, Mills, & Roseborough, 1951).
Accordingly, our conception and operationalization of hierarchical rank includes power and
status.
Principled Dissent
Graham (1986, p. 2) introduced the construct of principled dissent, defining it as any
effort individuals make to protest and/or change the organizational status quo because of their
conscientious objection to currently policy or practice. By the term “conscientious objection,”
she means, fundamentally, moral objection, as illustrated by her review of the moral judgment
literature and her claim that principled dissent arises in response to the perception of moral
wrongs in the workplace. In one sense, the original definition seems to imply that an actor’s
HIERARCHICAL RANK AND PRINCIPLED DISSENT
8
psychological state defines what is or is not principled dissent. However, Graham (1986, p. 2)
states that “the term principled applies to the issue at stake, e.g., one which violates a standard of
justice, honesty, or economy: it does not necessarily describe the ultimate motive of the person
who raises it.” To clarify this point but remain true to her conceptualization of the construct, we
define principled dissent as an individual’s effort to protest or change morally objectionable
practices.
Principled dissent is the first step toward improving ethical behavior in an organization
(Brief, Buttram, & Dukerich, 2001; Nemeth & Staw, 1989). It is a type of political action
(Cavanaugh, Moberg, & Velasquez, 1981; Farrell & Peterson, 1982) taken by people internal to
the organization. Relative to other approaches to correcting organizations’ ethical failures, such
as free market discipline or political pressure from others external to the organization
(Hirschman, 1970), principled dissent is a relatively efficient way to stop unethical practices. It is
important to study principled dissent because it helps to rectify ethical failures before they can
damage an organization.
According to Graham (1986, p. 3), principled dissent can take a variety of forms,
including constructive criticism or protest expressed to others within the organization, reports to
audiences outside the organization, blocking actions, and resignation accompanied by an
explanation. The defining element is the effort to protest a perceived moral wrong within the
organization (Graham, 1986; Miceli, & Near, 1992). Acts need not be effective in order to
qualify as principled dissent; many factors influence whether principled dissent is effective
(Detert, Burris, Harrison, & Martin, 2013; Miceli & Near, 2002; Near & Miceli, 1995).
HIERARCHICAL RANK AND PRINCIPLED DISSENT
9
The Importance of Principled Dissent by High-Ranking Individuals
Without principled dissent, unethical practices can become widely accepted by
organizational members. Because groups seek cohesion, people often conform in group contexts
(Asch, 1956; Festinger, 1950; Janis, 1972; Nemeth & Staw, 1989). Just as people must
sometimes revise their tasks to ensure optimal performance (Staw & Boettger, 1990), people
sometimes must engage in principled dissent to ensure group practices are ethical. It is especially
important for high-ranking people at the top of the organization to engage in principled dissent in
response to unethical practices (Ashforth & Anand, 2003; Brief et al., 2001; Mayer et al., 2009;
Mayer et al., 2013). From the perspective of others, those in high-ranking positions are more
competent, experienced, and committed to collective goals (Magee, Kilduff, & Heath, 2011).
Thus, people may trust those in high-ranking positions to provide direction, relying on their
judgments as heuristics for what is ethical (Strudler & Warren, 2001).
Additionally, lower-ranking individuals may not be well positioned to oppose unethical
behavior (Brief et al., 2001). Because low rank confers relatively little power, low-ranking
individuals could be constrained by fear of punishment, high levels of negative affect, and stress
(Anderson & Berdahl, 2002; Keltner, Gruenfeld, & Anderson, 2003; Sherman et al., 2012).
Similarly, low rank confers relatively little status, and lower-status individuals are in fact
punished more severely for norm violations (Becker, 1963; Bowles & Gelfand, 2010; Hollander,
1958, 1961). Thus, principled dissent could be risky for those with lower rank. Because their
positions confer power and status, higher-ranking individuals are arguably better equipped
psychologically and politically to engage in principled dissent.
Will High-Ranking Individuals Engage in Principled Dissent?
HIERARCHICAL RANK AND PRINCIPLED DISSENT
10
How likely is it that those at the top of the hierarchy will engage in principled dissent
against unethical practices in their organization? The psychological consequences of hierarchical
rank per se are not well known, but because hierarchical rank confers both power and status
(Fragale et al., 2012; Magee & Galinsky, 2008; Tost, 2015; Weber, 1948) we can draw from
those literatures to make predictions.
Why might holding higher rank increase principled dissent? Because hierarchical
rank confers power, holding higher rank could increase principled dissent when an ethical
problem is perceived. Power activates an approach orientation characterized by uninhibited
action and decreased sensitivity to threats (Anderson & Berdahl, 2002; Keltner et al., 2003). As a
result, powerful individuals are driven more by internal states than by external constraints
(Galinsky et al., 2008; Pitesa & Thau, 2013). For instance, when primed with power, communal
individuals act more generously whereas exchange-oriented individuals act more self-
interestedly (Chen, Lee-Chai, & Bargh, 2001). Similarly, DeCelles, DeRue, Margolis, and
Ceranic (2012) found that power is associated with self-interested behavior only for those with a
weak moral identity. Those with a strong moral identity act less self-interestedly following a
power prime. In a field study, Pitesa and Thau (2013) found that low- but not high-power
individuals conformed to organizational pressures to act ethically. Instead, high-power
individuals’ ethical action was driven by their moral identity strength. Together, this research
suggests that high-ranking individuals act in accord with their traits and values, and might
therefore engage in more principled dissent than lower-ranking individuals, when they perceive
an ethical problem.
Whether high-ranking individuals will perceive an ethical problem when one exists is
unclear, however. To date, this issue has not been explored empirically. Graham (1986) proposed
HIERARCHICAL RANK AND PRINCIPLED DISSENT
11
a positive relation between level in the organization and principled dissent because she expected
higher-ranking individuals to have greater awareness of ethical problems in the organization and
feel greater responsibility for attending to them. Our predictions differ from her perspective
because we believe high-ranking individuals are less likely to recognize ethical problems that
have emerged. We elaborate on this position below.
Why might holding higher rank decrease principled dissent? The power conferred by
high-ranking positions could limit detection of ethical problems for a number of reasons.
Generally, power does not promote morally upstanding behavior. It promotes a host of
objectionable behaviors, including selfishness (Blader & Chen, 2012), objectification of other
people (Gruenfeld et al., 2008) and a lack of empathy (van Kleef et al., 2008) and perspective-
taking (Galinsky, Magee, Inesi, & Gruenfeld, 2006), all of which could prevent high-ranking
individuals from seeing unethical practices as such. In the current research, we propose another
reason that is perhaps more subtle but that has important implications. Specifically, we
hypothesize that holding higher rank in a hierarchy will lead individuals to identify with their
group more, which in turn will lead to seeing its practices as more ethical. Together, these
mechanisms could prevent high-ranking individuals from perceiving ethical problems and
engaging in principled dissent.
High rank and group identification. An important feature of higher rank in
organizational hierarchies is that it is almost always accorded by the organization. Individuals are
not randomly assigned to positions of higher rank, nor are they typically foisted upon the
organization by some external party. Rather, individuals are promoted to high-ranking positions
by their organization – specifically, by a committee, a supervisor, a human resources department,
or other party within their organization.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
12
Current research has not closely attended to this aspect of high-ranking positions. Studies
on the related constructs of power and status often asks individuals to recall a time in which they
did or did not have power or status (Galinsky, Gruenfeld, & Magee, 2003). This method has
many benefits. It strips away many confounding variables that co-occur with power and status in
the real world, encouraging scientific rigor. It has also allowed for prolific study of important and
pervasive variables (Magee & Galinsky, 2008). However, this manipulation involves asking
individuals to recall a time in which they held power in a context separate from the one in which
their behavior is assessed (i.e., in the past even when their behavior is observed in the
laboratory). For instance, in one study participants were randomly assigned to feel powerful or
powerless by recalling a time they had power over someone or someone had power over them;
then, they reacted to participants’ attitudes in the current study session by cheating in a dice-
rolling game (Lammers, Stapel, & Galinsky, 2010). Consequently, this method does not allow for
exploring the impact of high-ranking positions on identification with the groups and
organizations that accorded them, and identification as a construct has been largely neglected by
the hierarchy literature to date.
After being accorded a high-ranking position in a group, individuals are likely to view
their group more positively and feel more identified with it in return. Identification is “the degree
to which people cognitively merge their sense of self and their evaluations of self-worth with
their judgments of the characteristics and status of their group” (Tyler & Blader, 2003, p. 354).
Why would higher rank elevate identification? One reason is that, by according high-ranking
positions, groups and organizations meet members’ psychological needs, thereby becoming more
attractive to those who hold high-ranking positions than to those do not hold such positions. In
addition to enhancing feelings of control (Fast, Gruenfeld, Sivanathan, & Galinsky, 2009), the
HIERARCHICAL RANK AND PRINCIPLED DISSENT
13
power that accompanies high-ranking positions leads to positive emotion (Keltner et al., 2003),
subjective well-being (Kifer, Heller, Perunovic, & Galinsky, 2013), and lower stress (Carney,
Cuddy, & Yap, 2010). Additionally, power increases self-esteem (Wojciszke, Struzynska-
Kujalowicz, 2007). Similarly, higher social status confers a host of benefits including greater
self-esteem, subjective well-being, and physical health (Anderson, Hildreth, & Howland, 2015).
Accordingly, the desire for status is a fundamental and powerful motive (Tay & Diener, 2011).
Holding a higher-ranking position therefore confers psychological and material advantages,
making the group more attractive to high- than low-ranking individuals, and people identify with
attractive groups that meet their psychological needs (Dutton, Dukerich, & Harquail, 1994).
Consistent with our proposition, some prior work has found positive relations between
identification and holding a higher-ranking position. For example, Kreiner and Ashforth (2004)
found a positive relation between being a supervisor and organizational identification, but the
causal direction was unclear. The authors believed identification could help individuals achieve
promotions to higher-ranked positions. Additionally, Willer (2009) found that people who were
accorded higher social status identified more strongly with their group. Similarly, research has
found that people identify more strongly with powerful roles, relative to less powerful ones
(Joshi & Fast, 2013). Importantly, the high-ranking roles in those studies were ones in which the
person was accorded higher rank or power in a specific group or organization, and then
expressed identification with that specific group or organization.
Identification and principled dissent. If holding higher rank increases identification,
what are the implications for principled dissent? At first glance, the answer is not obvious
because dissent can be helpful to groups. Although dissent compromises shorter-term group
stability goals, it promotes longer-term group change goals (Packer, Fujita, & Chasteen, 2014;
HIERARCHICAL RANK AND PRINCIPLED DISSENT
14
Packer, Fujita, & Herman, 2013). Highly identified group members often dissent when they
perceive doing so to help the group (Packer, 2009; Packer & Chasteen, 2010). Why, then, would
they fail to dissent against unethical practices?
We expect high levels of identification to prevent high-ranking individuals from
perceiving ethical problems. That is, high-ranking individuals could be less likely to experience
“normative conflict,” or disagreement with group norms (Packer, 2009) in the first place. People
are motivated to maintain positive views of the groups with which they identify (Dutton et al.,
1994; Ellemers, Spears, & Doosje, 2002), and morality is an important dimension on which
groups are evaluated (Leach, Ellemers, & Barretto, 2007). Consequently, identification is
associated with more positive perceptions of the group’s morality (Leach et al., 2007). It is
difficult for high identifiers to view their group as unethical because doing so puts them out of
balance (Heider, 1958). Viewing group norms negatively means they must also view themselves
negatively (Ellemers et al., 2002). By this logic, identifying highly with a group will motivate
people to see group practices as ethical, and to accept these practices.
To the extent that high-ranking individuals who identify highly with their group or
organization view practices as more ethical than do lower-ranking individuals, they will lack
motivation to engage in principled dissent. Experiencing normative conflict, or disagreement
with group norms, is one necessary pre-condition for high-identifiers to express dissent (Packer
& Miners, 2014). Notably, this is not to say that holding higher rank compromises individual
agency. Highly identified group members make decisions on the basis of their own judgments
about the interests of the group (Reicher, Haslam, & Smith, 2012). Holding higher rank simply
leads to more positive views of the group’s decisions and practices. Then, high-ranking
individuals act on their own (more positive) ethical views. In line with this logic, Dukerich et al.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
15
(1998) theorized that people who identify strongly could defer too readily to other group
members and fail to question the ethicality of group actions.
Therefore, if individuals in high-ranking positions identify highly with their organization
or group, they may see less need for principled dissent because they do not perceive their
organization’s or group’s practice as unethical in the first place. Perceiving a practice as unethical
is the first step toward principled dissent (Graham, 1986; Packer & Miners, 2014). Although
existing research suggests that the power that comes with higher rank provides greater license to
engage in principled dissent (Galinsky et al., 2008; Keltner et al., 2003; Pitesa & Thau, 2013),
higher-ranking positions might limit individuals’ ability to see when principled dissent is
necessary. Organizational leaders might not engage in principled dissent simply because they see
no need to do so.
Some empirical evidence provides preliminary support for these arguments. Although no
research to date has tested for causal relations between hierarchical rank, identification, and
principled dissent, some research shows an association between identification with a group and
stronger endorsement of its goals and values (Jetten, Postmes, & McAuliffe, 2002; Sechrist &
Young, 2011). In addition, people who identify highly with a group are more influenced by its
norms (Binning, Brick, Cohen, & Sherman, 2015; Hertel & Kerr, 2001; Mackie & Cooper, 1984;
Jetten et al., 2002; Zou, Morris, & Benet-Martinez, 2008). Similarly, high levels of identification
with a group or organization produce cooperation with its goals (Dukerich, Golden, & Shortell,
2002; McDonald & Westphal, 2011; O’Reilly & Chatman, 1986). Generally, this has positive
effects for organizations (Mael & Ashforth, 1992; Pratt, 2000) by leading to performance
improvements (Walumbwa et al., 2011), reducing agency costs, and requiring less external
HIERARCHICAL RANK AND PRINCIPLED DISSENT
16
governance controls (Boivie, Lange, McDonald, & Westphal, 2011). Nevertheless, weaker
principled dissent could be one cost of stronger identification.
Research has also found that those higher in the hierarchy view their organization more
positively in general (Stouffer et al., 1949; Tannenbaum et al., 1974). For instance, Trevino,
Weaver, and Brown (2008) found that senior managers saw the internal ethical environment of
their organizations more positively than did lower level employees. Similarly, Guimond (1995)
found that cadets ranking higher in a military hierarchy endorsed military values more strongly,
and Guimond, Dambrun, Michinov, and Duarte (2003) found that upper-level students adopted
their profession’s attitudes more than lower-level students. Although these studies were mainly
correlational and left the theoretical mechanisms unspecified, they provide some indirect
preliminary support for the idea that powerful, higher-ranking individuals accept their
organizations’ values.
One study has examined the effect of high rank on whistleblowing, a form of principled
dissent. Miceli and Near (1984) predicted that higher-ranking individuals would report unethical
activities more than lower-ranking individuals because their greater pay and education reduce
their dependence on the organization. However, their findings were equivocal, leaving this
question open for further exploration.
In sum, we expect that individuals in high-ranking positions will be less likely to engage
in principled dissent within their organization, in part because their identification leads them to
view existing practices within the group as being more ethical. Correspondingly, even when the
practices of a group or organization are unethical, individuals in high-ranking positions will be
unlikely to engage in principled dissent against these practices.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
17
Hypothesis 1. Individuals in high-ranking positions will be less likely to engage in
principled dissent, relative to those in lower-ranking positions.
Hypothesis 2. Greater identification among high-ranking individuals will partially
explain the negative relation between holding higher rank in the group and principled
dissent.
We acknowledge that there are many possible reasons why individuals at the top of the
hierarchy might fail to engage in principled dissent; for example, they might particularly benefit
from the unethical practice (e.g., a higher year-end bonus), they might be involved in the
unethical act themselves (e.g., by giving unethical orders), or they might face stiffer penalties if
the practice were to become publicly known (e.g., termination or jail time). We focus on
identification because we expect the effects of possessing high rank on identification to be robust
across group and organizational contexts. In contrast, factors such as whether high-ranking
individuals profit from unethical practices or whether they face stiffer penalties can vary more
from one context to another, and might thus be less generalizable.
Study 1: Principled Dissent in U.S. Federal Government Agencies
Study 1 examined archival survey data collected from over 11,000 employees of U.S.
federal government agencies. This data set was ideal for our purposes for many reasons. First, it
involved a very unique survey that asked an unusually broad range of employees whether they
engaged in whistleblowing, or had reported any unethical or even illegal activities by their
coworkers and colleagues. Whistleblowing is one type of principled dissent. It is defined as “the
disclosure by organizational members (former or current) of illegal, immoral, or illegitimate
practices under the control of their employers, to persons or organizations that may be able to
effect action” (Near & Miceli, 1985, p. 4). Like other forms of principled dissent, whistleblowing
HIERARCHICAL RANK AND PRINCIPLED DISSENT
18
protests unethical activity. Second, because it involved a very large sample size, it provided
ample statistical power. Third, holding a high-ranking position in these government agencies was
highly representative of different kinds of organizations and industries, in that it carried the role
requirements and responsibilities associated with high rank and control in real organizations.
Fourth, the survey was administered to over 22 different government agencies that were very
different from each other, further increasing the generalizability of any results. Fifth, ethical
transgressions and illegal activity in these agencies could have very serious consequences, and
could include behaviors ranging from stealing federal funds to accepting bribes to allowing a
practice that is dangerous to public safety. Sixth, many organizations have complex hierarchical
structures, making employees’ hierarchical rank difficult to measure, but rank in these
government agencies is indexed in a straightforward way. Finally, the survey was completed
privately and anonymously, which minimized pressure on participants to misreport whether they
had tried to engage in principled dissent in the past.
Sample
The U.S. Merit Systems Protections Board conducted a survey of 11,162 employees of
U.S. federal government agencies in 1992. The participants represented approximately a 54%
response rate among a total of 20,851 people who were randomly selected. Over 22 agencies
were represented, including NASA, the Environmental Protection Agency, the Defense
Department, the State Department, and the Department of Justice. The sample consisted of 58%
men and 40% women (2% of participants did not provide demographic information) and it
exhibited a diverse age distribution. Participants reported their age category. The mean was 4.11
(SD = 1.33), indicating the average age was approximately 40 years. The sample was 75%
HIERARCHICAL RANK AND PRINCIPLED DISSENT
19
Caucasian, 13% African American, 5% Hispanic, 3% Asian, 2% Native American, and 1% who
reported “other” backgrounds.
Measures
Hierarchical rank. Participants’ hierarchical rank was operationalized as level in the
agency, of which there were five coded in the survey. The first level included trade, craft, and
laboring employees; the second included supervisors of trade, craft, and laboring employees; the
third included professional, technical, and administrative employees; the fourth included
managers; and the fifth included executives. Dummy variables were created to represent rank,
with the first level serving as the reference group. We used dummy variables because rank may
not have been an interval scale, though as we report, measuring rank as a continuous variable did
not change the results.
Principled dissent. Respondents indicated whether they had personally reported illegal or
wasteful activity involving their agency to any of a number of parties within the last 12 months.
This served as a measure of whether they had engaged in principled dissent. Response options
are provided in Appendix A. We coded the reporting measure “1” if respondents reported the
activity to anyone and “0” if they did not report the activity. Overall, 16% of participants
reported such activity to at least one party.
Control variables. Individuals’ gender, ethnicity, organizational tenure, and age can
predict the individuals’ hierarchical rank and might also affect whether individuals engage in
principled dissent. To help rule out the possibility that these factors might act as third variables
and lead to a spurious relationship between power and reporting unethical practices, we
controlled for them in our analyses.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
20
Women tend to be less represented in high-ranking positions within organizational
hierarchies than men (Catalyst, 2013), and they tend to have higher, less flexible ethical
standards (Kray & Haselhuhn, 2012; Kennedy & Kray, 2013). Individuals of minority racial or
ethnic groups also tend to attain less high-ranking positions in groups and organizations (Berger,
Cohen, & Zelditch, 1972; Berger, Webster, Ridgeway, & Rosenholtz, 1986). Members of
marginalized, disadvantaged groups might be more likely to disagree with their groups’ practices
in general (Hogg, Fielding, & Darley, 2005). Therefore, we controlled for gender and ethnicity.
Similarly, education tends to correlate positively with hierarchical position (Carmeli,
Shalom, & Weisberg, 2007), in that people of lower education tend to occupy less high-ranking
positions. We included education as a control variable to rule out the possibility that less
educated employees might report practices as unethical out of feelings of marginalization.
Respondents reported their highest education level on a scale ranging from 1 (less than high
school diploma) to 7 (graduate or professional degree). The mean was 4.48 (SD = 1.68),
indicating that the average employee had achieved an associate’s degree.
Another concern was that tenure in the organization might drive a relation between
hierarchical rank and reporting unethical practices. Tenure and rank in the organization are often
positively correlated in organizations, as people move up the hierarchy the longer they work in
the organization (Carmeli et al., 2007), and individuals tend to adopt their organizations’ values
over time (Chatman, 1991). Therefore, we controlled for tenure. Respondents reported how
many years they had been a federal government employee (excluding military service) using a
scale ranging from 1 (less than 1 year) to 8 (31 years or more). The mean was 4.41 (SD = 1.82),
indicating the average tenure was approximately 13 years. In the same vein, age could be
positively correlated with higher rank. Prior research has found differences in ethical judgment
HIERARCHICAL RANK AND PRINCIPLED DISSENT
21
by age (for a review, see O’Fallon & Butterfield, 2005). Therefore, we controlled for age.
Respondents reported their age on a scale from 1 (under 20) to 8 (65 or older). The mean was
4.11 (SD = 1.33), corresponding to an average age of 40 to 49.
Finally, it was possible that hierarchical rank might correlate with knowledge of rules
about retaliation for reporting unethical practices; perhaps individuals lower or higher in the
hierarchy were exposed to more training, or more recent training, on this topic. Therefore we
controlled for knowledge about the rules of retaliation for reporting unethical practices. Using a
scale of 1 (a lot) to 4 (nothing), respondents reported how much they knew about the actions they
could take if they were retaliated against for reporting illegal or wasteful activity. We reverse-
scored this measure so that higher numbers indicated greater knowledge, M = 2.08, SD = 0.96.
Results
Table 1 provides descriptive statistics and correlations among all variables.
In a logistic regression analysis, we predicted principled dissent with our control
variables – gender, ethnicity, education, tenure, age, and knowledge of protections from
retaliation – as well as our primary independent variable, employees’ hierarchical rank. The
results appear in Table 2.
Presented in Model 1 of the table are relations between the control variables and
principled dissent. As shown, gender significantly predicted principled dissent, with women
engaging in less principled dissent, rather than more, than men, z = -2.46, p = .01. Similarly,
those of non-Caucasian ethnicity were less, not more, likely to report such activity, z = -4.02, p
< .001. Principled dissent was positively related to education level, z = 2.28, p = .02. This
suggests more educated individuals were more likely to engage in principled dissent, perhaps
because education hones individuals’ ability to think critically and independently (Tsui, 1999),
HIERARCHICAL RANK AND PRINCIPLED DISSENT
22
giving them the skills needed to detect when unethical activity is occurring. Tenure in the
organization was not related to principled dissent, z = 0.02, p = .99. Age related negatively to
principled dissent, z = -4.02, p < .001, suggesting individuals who were older were less likely to
report unethical practices. Finally, knowledge of protections from retaliation predicted principled
dissent, z = 2.52, p = .01. Individuals were more likely to engage in principled dissent when they
had more knowledge of the protections against retaliation.
We next examined the overall effect of hierarchical rank on principled dissent, above and
beyond the control variables. As shown in Model 2 of Table 2, a Wald test showed a significant
omnibus effect of employees’ hierarchical rank, as a set of variables, on principled dissent. This
suggests that principled dissent had a robust negative relationship with holding higher rank in the
hierarchy. We also examined the relation between each step up in the power of employees’
position and principled dissent, as shown in Model 2 of Table 2. Each step up in the hierarchy
was associated with a lower likelihood of principled dissent (the step from Level 1 to Level 2
being marginally significant). Indeed, the association between employees’ hierarchical rank and
principled dissent was strong: Relative to employees in the lowest ranking positions, executives
had approximately 64 % lower odds of principled dissent, z = -4.61, p < .001. Figure 1 displays
the frequency of principled dissent as a function of rank. As a robustness check, we conducted
the same analysis using hierarchical rank as a continuous variable. Again, the odds of reporting
unethical practices were lower as rank increased, OR = 0.78, z = -5.94, p < .001.1 Together, these
1 As additional robustness checks, we tested the relation when principled dissent was operationalized differently.
Including only reports to people within the organization (excluding reports to family and friends), holding higher
rank still predicted lower likelihood of principled dissent, χ (4) = 17.56, p = .002. The odds of principled dissent
were 47% lower for the highest-ranking employees relative to the lowest-ranking ones, OR = 0.53, z = -2.43, p = .
02. Including only reports to superiors at work, results were again very similar, χ (4) = 14.41, p = .006, OR = 0.58, z
= -1.94, p = .05. To account for culture differences across agencies, we also ran the analysis with 21 dummy
variables for government agency. Results were virtually identical. Hierarchical rank had a significant negative
relation with principled dissent, χ (4) = 32.48, p < .001, OR = 0.38, z = -4.29, p < .001.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
23
effects provided support for Hypothesis 1 within a field setting. Holding a higher-ranking
position in the government agencies was associated with a lower likelihood of reporting
unethical practices.
Discussion
Study 1 found a significant negative relation between employees’ hierarchical rank and
the likelihood that they would engage in principled dissent, supporting Hypothesis 1. Employees
in high-ranking positions were less likely to report unethical practices than those in low-ranking
positions. This relationship held up even after controlling for a host of variables that also had a
significant relation with principled dissent, including gender, ethnicity, education, tenure with the
organization, age, and knowledge of protections from retaliation. The findings thus lend
confidence that the relation between hierarchical rank and principled dissent was not due to any
of those possible third variables. Moreover, the link between hierarchical rank and principled
dissent was strong, suggesting the highest-ranking employees had 64% lower odds of principled
dissent than those in the lowest rank.
Despite these strengths, the study did not allow us to address the direction of the effect or
other possible alternative explanations. For example, it is possible that employees in more high-
ranking positions were less likely to report unethical activity because they were more likely to
take corrective action to fix the misconduct, rather than report it to others. Alternatively, people
at different ranks may have perceived different types or frequencies of unethical practices. We
thus designed Study 2, a controlled laboratory experiment, to address the limitations of Study 1.
In Study 2 and thereafter, we explore identification as one possible mechanism underlying the
relation between hierarchical rank and principled dissent. Our studies cannot ensure that
identification is the only or most important mechanism in Study 1, however. Here, those with
HIERARCHICAL RANK AND PRINCIPLED DISSENT
24
high rank might have particularly benefit from the unethical practice (e.g., a higher year-end
bonus) or been involved in the unethical act themselves (e.g., by giving unethical orders), for
instance. Our goal in future studies is simply to test identification as one possible mechanism.
Study 2: The Manipulation of High-Ranking Positions in Laboratory Groups
We had four primary aims in Study 2. First, to establish causality, we used an
experimental design. Participants were randomly assigned to positions of high or low rank within
a group. A key feature of the rank manipulation was that rank was ostensibly afforded to
individuals by their laboratory group, similar to how individuals in organizations are typically
afforded a high-ranking position by others in that organization – for example, a supervisor or
personnel committee – rather than some external party.
Second, Study 1 aimed to test the effects of possessing high vs. low rank separately.
Holding a high-ranking position might have not necessarily have the opposite effect of holding a
low-ranking position. For example, holding a high-ranking position might decrease principled
dissent (relative to a neutral position), holding a low-ranking position might increase principled
dissent (relative to a neutral position), or both could occur.
Third, the study tested our hypothesized mechanism: identification. We expected
identification to explain the negative relation between holding a high-ranking position and
principled dissent.
Finally, this study sought to rule out an alternative explanation for the findings in Study
1. We have argued that holding a high-ranking position leads individuals to identify more with
their organization or group, which in turn leads them to engage in less principled dissent.
However, it is possible that higher rank was associated with less principled dissent in Study 1
simply because holding higher rank is associated with power, which corrupts (Kipnis, 1972).
HIERARCHICAL RANK AND PRINCIPLED DISSENT
25
Prior research has found that the possessing power can promote unethical behavior (Georgesen &
Harris, 1998; Gruenfeld, Inesi, Magee, & Galinsky, 2008; Kipnis, 1972). Thus, it is possible
individuals in high-ranking positions might have engaged in less principled dissent simply due to
a predilection to act unethically, not because of higher identification with their organization and
subsequent acceptance of their organization’s unethical practices. Said differently, high rank
might have produced a more unethical disposition.
To test this alternative explanation, Study 2 experimentally manipulated whether the
group recommended an unethical or ethical course of action. If higher rank leads to less dissent
in both conditions, it would suggest that higher rank leads to greater acceptance of group
practices in general, whether those practices are ethical or unethical. In contrast, if higher rank
led individuals to dissent less only when the group sets an unethical practice, but dissent more
when the group sets an ethical practice, it would suggest that higher rank simply corrupts and
leads people to prefer unethical actions.
Sample
Participants included 271 adults (67% women) affiliated with a West Coast university as
undergraduate or graduate students, staff, alumni, or local community members. They received
$15 for participating. The sample was 22% Caucasian, 60% Asian, 7% Hispanic, 2% African
American, and 6% who reported other backgrounds. Their ages ranged from 18 to 63 years (M =
21.55, SD = 6.17). Five (2%) participants were suspicious about the veracity of the manipulation
and were excluded from analyses. Results were virtually identical when they were included.
Design and Procedure
The experiment had a 3 (Hierarchical rank: High, low, control) x 2 (Group morality:
High, low), between-participants design. Participants reported to the laboratory in groups of six
HIERARCHICAL RANK AND PRINCIPLED DISSENT
26
for a 60-minute session consisting entirely of this experiment. There were 47 groups who
interacted together for one hour. If only five participants showed up for a given session, a
research assistant stood in as a confederate.2
The experiment had three parts: an icebreaker exercise, selection of a high-ranking group
member, and a group task that provided the opportunity for dissent. First, participants introduced
themselves to their group by stating their first name, field of study or a recent job, and favorite
thing about their city of residence. Later, in the group task, participants would believe they were
making a decision that affected another group in a separate laboratory space. To establish realism
for the upcoming inter-group exercise, the experimenter called a research assistant to ask
whether the “other group” was full and if they were ready to begin the study. In fact, there was
no other group in the other laboratory space.
Using a method adapted from previous research (Leary, Cottrell, & Phillips, 2001; Marr
& Thau, 2014), the second part of the study involved ostensibly selecting a group member to
hold a high-ranking position in the group. Participants were separated into different computer
terminals and they completed a personal information questionnaire based on the one used in
Leary et al. (2001), which measured personality characteristics, personal behaviors and habits,
and political and moral attitudes. Then, to further establish the legitimacy of the peer-
nominations for rank, participants read a short Harvard Business School case study (Hamermesh,
Whittemore, & Sherman, 2010) and reported in three sentences how they would recommend the
protagonist handle the situation described in the case. Participants were informed in advance that
the group would see their answers to these questions. As participants submitted their answers,
2 A research assistant filled in for a participant in 11 out of 47 groups. Excluding these 11 groups, results were
virtually identical. High rank reduced dissent relative to the control condition, OR = 0.20, z = 8.27, p = .004.
Similarly, when a dummy variable is entered to account for the presence of a confederate, this variable is non-
significant, z = 0.65, p = .42. Together with our very low rates of suspicion (2%), these analyses give us confidence
that our occasional use of confederates in place of group members did not impact our results.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
27
their computers showed a dialogue box saying that the group members’ answers were printing.
After all participants had ostensibly printed their answers, a research assistant delivered a packet
of responses to the experimenter. Participants believed these responses were from their fellow
group members, but they were actually pre-scripted.
These materials appeared to serve as the basis for the selection of the high-ranking group
member. Participants read that, because decision-making often occurs in hierarchical contexts,
they would choose someone to serve as leader of the group. This high-ranking member was
described as responsible for making key decisions, overseeing the group’s performance,
evaluating other group members, and determining whether each member should receive a bonus
for their participation (see also Anderson & Berdahl, 2002). To select this person, participants
reviewed the printed materials ostensibly from each fellow group member and allocated 10
points across the 5 other members of the group. They could divide these points among the other
participants however they wished and were led to believe that the person with the most points
would be selected as the group’s high-ranking member.
Manipulation of high-ranking positions. After a randomly generated wait time between
1 and 2 minutes, participants received one of three randomly determined messages. In the high
rank condition, they read that the group allocated them 32 points and had chosen them to serve as
the high-ranking member of the group. In the low-rank condition, participants read the group had
allocated them 6 points and did not select them to be the high-ranking member. In the control
condition, participants received no information regarding the distribution of points. They simply
read that results would be announced later. Participants then answered a short survey with
measures of identification and the manipulation check.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
28
Ethical decision-making task. In the third and final part of the study, participants
engaged in the Deception Game (Gneezy, 2005), in which their group would decide whether to
lie to another group for monetary gain. Specifically, participants were told that they would
participate in a task with a group down the hall. They read that two possible monetary payments
were available to their group and the group down the hall: Option A, which paid the participants’
group $90 and the other group $84, or Option B, which paid the participants’ group $84 and the
other group $90. The decision they had to make was whether to lie to the other group by sending
one of two messages. Message 1 would tell the other group, “Option A will earn you more
money than Option B,” while Message 2 would tell the other group, “Option B will earn you
more money than Option A.” Message 1 was false, and Message 2 was true. To enhance the
moral significance of this decision and to ensure that participants expressed genuine opinions
regarding what the group should do (not, for instance, a desire to tell the truth because they
thought the group down the hall would disbelieve their recommendation), participants also
received a message that indicated the other group had already committed to follow their
recommendation (see Cohen et al., 2009).
Separated from the rest of their group via cubicles, participants were told that their
group’s task was to decide, via an online chat-room, whether to send Message 1 (a lie) or
Message 2 (the truth). Before viewing the chat-room, however, all participants received a
message indicating they had been randomly assigned to report their recommendation to the
group in the fifth slot, or after four of the other group members had submitted their
recommendations. Therefore, because the group ostensibly had six members, participants
believed they were reporting their recommendation second to last. Based on Willer’s (2009)
HIERARCHICAL RANK AND PRINCIPLED DISSENT
29
procedure, this helped to minimize suspicion. Over the next few minutes, participants saw four
messages appear on the chat-room screen, purportedly sent from the other group members.
In the high group morality condition, the four messages indicated a unanimous preference
to tell the other group the truth, forgoing $6 for the group. In the low group morality condition,
the four messages indicated a unanimous preference to tell the other group members a lie,
gaining $6 for the group. These messages are included in Appendix B.
Providing the key dependent measure, participants selected which message they wanted
to send to the other group (Message 1 or 2). They also typed a short message to their own group
regarding their individual decision. Their short explanations appeared in the chat-room. After all
six group members had ostensibly reported their recommendations, the computer screen showed
a message that summarized the group decision. Before leaving, participants completed a short
survey to check their suspicion. Finally, participants were paid $15 (regardless of their decision),
thanked for participating, and informed that additional information about the study would come
by email. After data collection was complete, all participants were debriefed.
Measures
Hierarchical rank manipulation check. The manipulation check consisted of five items.
Participants rated the extent to which they had control over important resources in the group as
well as felt powerful, respected, admired, and high status in the group, on a scale of 1 (strongly
disagree) to 7 (strongly agree). The five items correlated highly and were averaged to form a
reliable scale, α = .87, M = 3.99, SD = 1.02.
Group identification. We used four items from prior research (Packer & Chasteen, 2010;
Willer, 2009) to measure group identification. Participants reported how much they identified
with the group, how much they felt connected with the group, and how much they valued being a
HIERARCHICAL RANK AND PRINCIPLED DISSENT
30
member of the group on a scale of 1 (not at all) to 6 (a great deal). They also indicated their
agreement that the group’s successes were their successes on a scale of 1 (strongly disagree) to 7
(strongly agree). The items correlated highly, so they were standardized and averaged to form a
scale, α = .80.
Dissent. We gave participants a “1” when they disagreed with their group by choosing to
send a message different from the one their group ostensibly wanted to send and a “0” when their
choice agreed with their group. In the low group morality condition, the measure captured
principled dissent (i.e., objection to an unethical practice). However, in the high group morality
condition, it measured unprincipled dissent (i.e., objection to an ethical practice). We therefore
refer to the measure as dissent, broadly. Across all conditions, 19% of participants opted to
dissent against their group’s decision on average.
We also measured the strength of participants’ dissent. Two independent coders blind to
the study conditions and hypotheses rated the strength of disagreement expressed in the
messages participants sent to their group members in the online chat-room. They rated the
messages on a scale of 1 (strongly agreed with the group) to 6 (strongly disagreed with the
group). For instance, one participant in the high group morality condition strongly agreed with
the group by writing, “I agree with the others, they trusted us and we should be honest.” This
received an average score of “1” indicating strong agreement. Another participant in the high
group morality condition disagreed by writing, “We want to max our earnings and we should do
whatever in our control that may favor that goal. So 1 [the deceitful choice].” This received an
average score of 5.5, indicating strong disagreement. The two sets of ratings correlated highly, r
(264) = .86, p < .001, and were averaged to form a continuous measure of strength of dissent, α =
.92, M = 2.48, SD = 1.32.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
31
Positive affect. One concern was that even if we found an effect of rank on dissent, it
might be due to positive affect. Receiving feedback that one has been chosen by the group to
hold rank might lead to positive emotion; in turn, it is possible that positive emotion might lead
to more general satisfaction with the group and therefore, less desire to disagree with the group’s
practices. To help rule out this alternative explanation we included the positive affect scale of the
Positive and Negative Affect Scales (PANAS; Watson, Clark, & Tellegen, 1988), using a scale of
1 (very slightly or not at all) to 5 (extremely), α = .90, M = 2.76, SD = 0.82.
Results
Table 3 provides zero-order correlations among all variables.
Manipulation check. The manipulation of rank was effective, F (2, 263) = 42.96, p < .
001, η2 = .25. Participants assigned to the high-ranking position (M = 4.67, SD = 0.94) perceived
themselves to have higher rank than those in the control (M = 3.90, SD = 0.75), t (173) = 6.01, p
< .001, d = 0.91, and low-ranking (M = 3.44, SD = 0.97), t (175) = 8.59, p < .001, d = 1.29,
conditions. Participants in the low rank condition perceived themselves to have lower rank than
those in the control condition, t (178) = 3.57, p < .001, d = 0.53.
Test of Hypothesis 1. Hypothesis 1 posited a negative effect of rank on principled dissent.
In a logistic regression, we predicted the odds of dissenting with hierarchical rank (using two
dummy variables) and group morality (high or low morality). The control condition served as the
reference group for rank, and the low morality condition served as the reference group for group
morality. The frequency of dissent by condition appears in Figure 2 and the logistic regression
results appear in Table 4.
Overall, those who were given a position of high rank had 75% lower odds of dissent
than individuals in the control (z = 8.52, p = .004) and low rank (z = 8.15, p = .004) conditions.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
32
Low rank did not affect the likelihood of dissent relative to the control condition, z = 0.08, p = .
93. Therefore, as predicted, individuals with high rank were less likely to dissent. Being
accorded low rank did not increase the likelihood of dissent, however.
A main effect of the group morality condition also emerged. Individuals in the high group
morality condition, in which the other group members ostensibly voted to tell the other group the
truth, had 82% lower odds of dissent than those in the low group morality condition, in which the
other group members ostensibly voted to lie to the other group, z = 20.43, p < .001. Therefore,
participants on the whole were more likely to accept their group’s decision if that decision was
honest. As shown in Model 2 of Table 4, no interactions between rank and group morality
emerged (ps > .68).
These results suggest that high-ranking individuals were less likely to dissent in general,
regardless of the group majority’s decision; that is, high-ranking individuals accepted the group
majority’s decision more than others regardless of whether the group members opted for an
unethical or ethical decision.
We then tested for gender effects, as prior research has found evidence that ethical views
differ for women and men (Franke, Crown, & Spake, 1997; Kennedy & Kray, 2013; Kray &
Haselhuhn, 2012). For dissent, a marginally significant main effect gender emerged, such that
women showed 45% lower odds of dissent than men, overall, z = 2.74, p = .098. No interactions
with gender reached significance (ps > .12). However, separate logistic regression analyses found
that women engaged in significantly less dissent than men in the high group morality condition
(OR = 0.21, z = 4.47, p = .03), but not in the low group morality condition (OR = 0.77, z = 0.38,
p = .54). That is, women were less likely to recommend lying if the group majority elected to be
honest. In light of these gender effects, we controlled for gender in subsequent analyses.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
33
We next analyzed the measure of dissent strength using ANOVA, as this dependent
variable was continuous rather than dichotomous. The findings were highly similar to those
above. Three main effects emerged. First, as expected, participants in the high rank condition (M
= 2.05, SD = 0.96) dissented less strongly than those in the control condition (M = 2.73, SD =
1.39), p = .001, F (2, 252) = 8.89, p < .001, ηp2 = .07. Once again, the main effect of rank on
dissent was driven by the effect of high rank. Strength of dissent did not vary between the low
rank (M = 2.78, SD = 1.45) and control conditions, p = .80. Second, there was a main effect for
group morality, F (1, 252) = 28.92, p < .001, ηp2 = .10. Participants dissented less strongly in the
high group morality condition (M = 2.01, SD = 1.03) relative to the low group morality condition
(M = 2.93, SD = 1.41), p < .001. No interaction emerged between rank and group morality, F (2,
252) = 0.41, p = .67. Third, a main effect emerged for gender, F (1, 252) = 4.63, p = .03, ηp2 = .
02. Women (M = 2.35, SD = 1.26) dissented less than men (M = 2.69, SD = 1.42). No
interactions with gender emerged (ps > .29). However, separate ANOVAs found that women (M
= 1.84, SD = 0.84) engaged in significantly less dissent than men (M = 2.35, SD = 1.29) in the
high group morality condition (F [1, 125] = 7.64, p = .007, ηp2 = .06), but not in the low group
morality condition (F [1, 127] = 0.45, p = .50, ηp2 = .004). Means on the strength of dissent
measure appear in Table 5.
Test of Hypothesis 2. Hypothesis 2 stated that identification with the group would
mediate the negative effect of rank on dissent. We tested this hypothesis first using the binary
measure of dissent. When identification (p = .02) was entered into the logistic regression, the
effect of high rank (p = .004) on dissent was reduced (p = .03) (see Model 5 of Table 4). A
bootstrapping analysis of mediation (Preacher & Hayes, 2008) with 10,000 re-samples with
replacement estimated the indirect effect to be -.33 and provided a 95% bias-corrected
HIERARCHICAL RANK AND PRINCIPLED DISSENT
34
confidence interval of [-.06, -.69]. This suggests that increased identification mediated the effect
of rank on dissent. Participants placed into the high-ranking position were less likely to dissent
because they identified more with their group
Next, focusing on the measure of dissent strength, we conducted a linear regression
analysis. The effect of high rank (b* = -.22, t [259] = -3.47, p = .001) was reduced (b* = -.17, t
[258] = -2.52, p = .01) when identification (b* = -.13, t [258] = -2.16, p = .03) was entered as a
predictor. The bootstrapping analysis estimated the indirect effect of identification to be -.15,
with a 95% bias-corrected confidence interval of [-.02, -.32]. This interval excluded zero,
suggesting that increased identification also mediated the effect of rank on the strength of
dissent. In other words, participants placed into the high-ranking position were not only less
likely to dissent but also dissented less strongly because they identified more with their group.3
Positive affect. One alternative explanation for our results is that placing individuals into
high-ranking positions might have simply made them happier and thus, less likely to disagree
with anyone, including their group. Consistent with existing theory, being placed into a high rank
position increased positive affect, F (2, 260) = 40.18, p < .001, ηp2 = .24. High-ranking
participants (M = 3.32, SD = 0.80) felt more positive affect than those in the control condition
(M = 2.46, SD = 0.74, p < .001). Low-ranking (M = 2.52, SD = 0.63) and control condition
participants did not differ in their positive affect, p = .57.
However, positive affect did not explain the effect of rank on dissent. We predicted
dissent with rank, group morality, gender, and positive affect in a logistic regression. Rank
predicted less dissent, OR = 0.29, z = 5.77, p = .02, but positive affect did not, z = 1.12, p = .29.
3 We also conducted mediation analyses excluding gender as a control. The effect of high rank (p = .004) on dissent
remained significant (OR = 0.31, z = 5.50, p = .01) when identification (OR = 0.65, z = 3.63, p = .057) entered the
logistic regression analysis. However, the bootstrapping interval bridged zero [-.64, .02]. For strength of dissent,
identification (b* = -.11, t [261] = -1.71, p = .088) again somewhat reduced the effect of high rank (p < .001) on
dissent (p = .01), but the bootstrapping interval bridged zero [-.28, .02]
HIERARCHICAL RANK AND PRINCIPLED DISSENT
35
Predicting strength of dissent with rank, group morality, gender, and positive affect in a linear
regression, higher rank again predicted less dissent, b* = -.19, t (258) = -2.64, p = .01, but
positive affect did not, t (258) = -1.07, p = .29. Thus, increased positive affect did not explain
why holding rank reduced dissent.
Discussion
Study 2 provided further support for Hypothesis 1. Holding a position of higher rank in
the group reduced the likelihood of disagreeing with the group’s unethical decision. It
complemented the field setting in Study 1 by using a controlled laboratory setting. The
laboratory setting also enabled us to test for a causal relation between rank and dissent.
This effect occurred across two courses of action recommended by the group – one
ethical, another unethical. Therefore, high-ranking individuals were less likely to disagree with
practices chosen by their group, regardless of whether the practices were unethical or ethical.
This helps rule out the alternative explanation that high rank simply corrupted, leading
individuals to opt for the unethical decision, regardless of the group’s preference. Rather,
individuals holding high rank accepted the group’s decision be it ethical or unethical. This
finding also suggests that high rank might lead to other forms of conformity beyond the domains
we examine here (a topic we return to in the General Discussion).
Low rank did not affect dissent relative to the control condition. This suggests that the
link between rank and dissent is due to high-ranking individuals dissenting less as opposed to
low-ranking individuals dissenting more. This asymmetric effect of high rank (but not low rank)
is consistent with prior research (e.g., Anderson & Galinsky, 2006; DeWall, Baumeister, Mead,
& Vohs, 2011; Goldstein & Hays, 2011; Mast, Jonas, & Hall, 2009).
HIERARCHICAL RANK AND PRINCIPLED DISSENT
36
Study 2 also provided support for Hypothesis 2, suggesting that an increase in
identification with the group among high-ranking individuals helped to explain why they
dissented less than individuals who did not hold high-ranking positions.
Study 3: Manipulating the Conditions that Enable (or Impede) Identification
Study 2 found that holding a high-ranking position led to less principled dissent because
it caused individuals to identify more strongly with their group. Another way to examine
identification as a mechanism underlying the effects of hierarchical position is to manipulate
experimentally the conditions under which identification might emerge – that is, to show that the
negative effects of high rank on principled dissent disappear when rank no longer affects
identification with the group or organization.
To do so, we randomly assigned participants to conditions that would enable (or impede)
the effects of rank on identification. Specifically, we created conditions under which participants
would be likely (or unlikely) to identify with a laboratory group. We also randomly assigned
positions of high or low rank. Then, we assessed how participants reacted to an unethical
practice. We predicted that high-ranking individuals would engage in less principled dissent than
lower-ranking individuals in conditions where they might identify highly with the group, but not
in conditions where identification was unlikely no matter their rank. If supported, this hypothesis
could further establish identification as one important mechanism underlying the effects of
hierarchical position on principled dissent.
Study 3 had two additional aims. First, it tested another aspect of the theoretical model:
the impact of rank and identification on ethical views. We measured perceptions of ethicality,
expecting high-ranking individuals – in conditions where they might identify highly – to engage
in less principled dissent because they perceive the group’s actions as being more ethical (or at
HIERARCHICAL RANK AND PRINCIPLED DISSENT
37
least less unethical). Second, this study examined principled dissent in a new context: price
gouging, a practice related to fairness, rather than deception (Kahneman, Knetsch, & Thaler,
1986).
Sample
Participants (N = 170) were undergraduates at a private, Southeastern university. They
received course credit for participating. The sample contained 80 women and was 72%
Caucasian, 11% Asian, 10% African American, 3% Hispanic, 2% Indian, 1% Middle Eastern,
and 2% “other” backgrounds. The average age was 19.76 years (SD = 1.08). In two of the 35
groups, a participant foiled the study procedure.4 We excluded data for those two groups (n =10),
and for eight participants (5%) who expressed suspicion regarding the study procedure, resulting
in a final sample of 152 observations.
Design and Procedure
The experiment had a 2 (Hierarchical rank: High, Low) x 2 (Identification: Enabled,
Impeded), between-participants design. Participants reported to the laboratory in groups of five
people. There were 35 groups who interacted together for 45 minutes. If only four participants
showed up for a given session, a research assistant stood in as a confederate.5
The experiment had four parts: introductions, selection of a high-ranking group member,
a manipulation of conditions that would enable or impede group identification, and a group task
that provided the basis for principled dissent. First, the experimenter gave the group five minutes
to prepare for introductions. To introduce themselves, participants stated their first name,
4 In one group, a participant looked at another person’s computer screen during the chatroom portion of the study. In
the other group, participants discussed their hierarchical rank feedback aloud.
5 A research assistant filled in for a participant in 3 out of 35 groups. Excluding these 3 groups, results were virtually
identical. When identification with the group was enabled, holding higher rank reduced principled dissent, OR =
0.29, z = 5.39, p = .02. When identification with the group was impeded, the relation between rank and principled
dissent was non-significant, OR = 1.44, z = 0.46, p = .50.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
38
hometown, major, favorite thing about their city of residence, biggest achievement to date, how
they would describe themselves in three words, and how often they read the Wall Street Journal.
This information served as an ostensive basis for the leader selection procedure that followed.
Then, participants completed the leader selection procedure from Study 2 by privately allocating
10 points among the four other members of the group. The experimenter collected the surveys
and appeared to tally up the points allocated to each person.
Manipulation of rank. Participants were randomly assigned to receive one of two
messages. In the high-ranking condition, participants read that the group allocated them 27
points and had chosen them to hold a high-ranking position in the group. In the low-ranking
condition, participants read the group had allocated them 5 points and they would hold a low-
ranking position in the group.
Manipulation of conditions enabling or impeding identification. Participants were then
randomly assigned to complete a manipulation drawn from prior research (Greenaway et al.,
2015) which varied how likely they were to identify with the laboratory group. In both
conditions, participants read a set of five negative statements and a set of five positive about their
school. After each set, they tallied how many of the negative and positive statements they agreed
with. Greenaway et al. (2015) found that tallying agreement with these positive and negative
items helped shape participants’ views of how strongly they identified with their school. In one
condition, the negative statements were difficult to agree with (e.g., I feel no real affiliation with
[school]; There is no sense of community spirit at [school]), and the positive statements were
easy to agree with (e.g., In general, I like attending [school]; At [school], there are many
recreational activities in which people can become involved), leading participants to believe they
identified highly with their school. In the other condition, the negative statements were easy to
HIERARCHICAL RANK AND PRINCIPLED DISSENT
39
agree with (e.g., I think it is good to have some friends who don't attend [school]; There are some
things I don’t like about [school].), and the positive statements were difficult to agree with (e.g.,
After I graduate, it will be essential for me that all my friends went to [school]; I don't
understand people wanting to attend universities other than [school]), leading participants to feel
they did not identify very highly with their school.
After everyone had completed the survey, the experimenter collected the surveys and
appeared to tally up responses. Then, the experimenter distributed a sheet that appeared to
summarize the group’s opinions. Across all conditions, participants received information that the
other group members identified highly with the school (i.e., strongly agreed with the positive
statements and disagreed with the negative statements, on average). Therefore, we expected that
feeling identified with their school would enable participants to identify with their group
members more easily (to a varying degree, depending on rank), whereas not identifying with
their school would impede identification with the other group members (no matter the
participant’s rank).
Ethical decision-making task. Participants then sat at computers separated by dividers,
ostensibly to complete a business decision-making task as a virtual group. Participants wore
headphones playing white noise in order to block out others’ typing patterns, which could create
suspicion during the chat-room portion of the study. They imagined themselves as employees at
Plasma International, a company in the business of supplying blood to hospitals. The scenario
was based on a business school case study of ethics (Zimmerer & Preston, 1976). After reading
some background information on the company, participants read that Plasma faced the following
business problem:
HIERARCHICAL RANK AND PRINCIPLED DISSENT
40
Plasma’s typical fee to blood donors is $5/pint—that is, Plasma pays people $5/pint for
their blood. After having the blood tested for disease and sanitized, Plasma typically re-
sells the blood to hospitals for $50/pint. Recently, there was a hurricane in the southern
United States that resulted in a great deal of injuries. As a result, demand for blood is
unusually high. Hospitals need more pints of blood than usual. However, supply of blood
has not increased enough to meet this need. There are not many more donors than usual.
Plasma International must decide whether to raise the price of blood to $100/pint or
whether to continue selling the blood at $50/pint. This will be the price charged to
hospitals.
Participants then read a short description of the benefits to each possible approach:
Each approach has its advantages. If Plasma sells the blood at $100/pint, demand for
blood will decrease to the point that Plasma will be able to fulfill all the orders that are
placed. Plasma will not have to make difficult and arbitrary decisions about who should
get the blood. Those who are willing to pay the most will receive it. In addition, Plasma
will maximize returns to shareholders, making more profit than usual. However, it will be
very expensive for hospitals to treat people and the hurricane's many victims are not
wealthy. If Plasma sells the blood at $50/pint, it will be more affordable for hospitals and
hurricane victims. But Plasma will not be able to fulfill all the orders that are placed. In
addition, Plasma will not maximize returns to shareholders, making less profit than would
be possible.
Finally, participants were told their group’s task was to make this decision together, as an internal
committee at Plasma. Participants were asked to decide as a group, via an online chat-room,
whether or not to raise the price of blood. Before viewing the chat-room, participants all received
HIERARCHICAL RANK AND PRINCIPLED DISSENT
41
a message that they had been randomly assigned to report their recommendation to the group in
the fourth slot, after three other group members had made their recommendations. In all
conditions, the three messages indicated a unanimous preference to raise the price of blood.
Based on participants’ real responses in a separate study, the responses read: “[First response]
Our obligation is to our shareholders and to this company, by raising the price to $100/pint we
are on track to fulfill both of those obligations. So option 1; [Second response] 1 – I am in
agreement. It’s a higher priced without going too crazy on the consumer; [Third response] I say
we choose 1. By raising the price, demand would drop, and supply would be enough to meet the
demand.” (We included typos and grammatical errors in order to enhance realism.)
To provide a measure of principled dissent, participants selected which option they
preferred (Option 1: Raise the price to $100/pint or Option 2: Keep the price at $50/pint). They
also typed a short explanation, which appeared in the chat-room. After all five participants had
ostensibly reported their recommendations, the computer advanced to a survey measuring views
of the group’s decision, the rank manipulation check, demographic characteristics, and suspicion.
Participants were debriefed via email after data collection was complete.
Measures
Table 6 provides descriptive statistics and correlations among variables. Participants
responded using a scale from 1 (strongly disagree) to 7 (strongly agree), unless noted otherwise.
Rank manipulation check. The rank manipulation check consisted of seven items.
Participants rated the five items from Study 2 and two additional items: whether they had control
over the group’s decision and the extent to which they felt they were held in high regard in the
group. The seven items correlated highly and were averaged to form a reliable scale, α = .92.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
42
Identification manipulation check. The identification manipulation check consisted of
six items: the four items from Study 2 and the number of positive and negative (reversed)
statements participants agreed with regarding their school. Note that although we manipulated
whether participants were presented with either positive or negative statements, the number of
positive or negative statements they endorsed still varied across participants. The six items were
standardized and combined, α = .74.
Principled dissent. To provide a measure of principled dissent, participants indicated
whether they recommend raising the price of blood to $100/pint (Option 1) or keeping the price
of blood at $50/pint (Option 2). We coded this measure “1” if participants disagreed with their
group by selecting Option 2, and “0” otherwise. On average, 32% of participants dissented.
Strength of dissent. As in Study 2, two independent coders rated the strength of
participants’ principled dissent using a scale from 1 (strongly agreed) to 6 (strongly disagreed).
The set of ratings correlated highly, r (150) = .93, p < .001. We thus averaged the ratings to form
one measure, α = .96.
Ethical views. Participants reported their ethical views by indicating their agreement that
their group’s decision was ethical, their group’s decision was moral, their group is an ethical
group of people, and overall, their group has high ethical standards, α =.92. We embedded these
items among two filler items (e.g., “my group’s decision was rational”).
Results6
6 In contrast to Study 2, we did not find an effect for gender on principled dissent (OR = 1.72, z = 2.24, p = .13).
Nonetheless, to be consistent with Study 2 we examined our results after controlling for gender. The results are
similar to those excluding it. When identification was enabled, holding higher rank reduced principled dissent (OR =
0.31, z = 4.77, p = .03), and gender was non-significant, OR = 0.85, z = 0.10, p = .75. When identification was
impeded, no effect of rank emerged (OR = 1.25, z = 0.18, p = .67), and women engaged in more principled dissent
than men, OR = 3.64, z = 5.51, p = .02. When identification was enabled and ethical views were entered into the
logistic regression, more positive ethical views predicted less principled dissent (OR = 0.16, z = 17.17, p < .001),
and rank (p = .03) became non-significant (p = .89). In a bootstrapping analysis (Preacher & Hayes, 2008), ethical
views (-3.51, -0.52) still mediated the effect of high rank when identification was enabled.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
43
We conducted analyses using ANOVA with rank and the manipulation of identification
conditions as between-participant factors, unless otherwise indicated.
Rank manipulation check. We first tested that our rank manipulation was effective. A
main effect emerged for hierarchical rank, F (1, 148) = 54.31, p < .001, ηp2 = .27. Participants in
the high rank condition (M = 4.20, SD = 1.15) perceived themselves to have higher rank in the
group than did participants in the low-rank condition (M = 2.84, SD = 1.19), confirming that our
manipulation of hierarchical rank was effective.
Identification manipulation check.7 For identification, two main effects emerged. Under
the conditions enabling identification, it was stronger (M = 0.30, SD = 0.59) than under
conditions impeding it (M = -0.29, SD = 0.59), F (1, 148) = 39.01, p < .001, ηp2 = .21. In
addition, those who held a position of high (M = 0.10, SD = 0.63) rather than low (M = -0.09, SD
= 0.67) rank identified more strongly, F (1, 148) = 4.27, p = .04, ηp2 = .03. Most important,
however, were the simple effects of rank within each of the identification conditions. As
expected, high-ranking participants identified more with their group (M = 0.44, SD = 0.51) than
low-ranking participants (M = 0.16, SD = 0.63) when identification was enabled, t (74) = 2.14, d
= 0.49. However, high-ranking participants did not identify with their group (M = -0.24, SD =
0.56) more than low-ranking participants (M = -0.35, SD = 0.62) when identification was
impeded, t (74) = 0.80, p = .43, d = 0.19. This suggests the manipulation was effective, in that in
7 We also conducted a pre-test measuring identification. Pre-test participants (n = 103) completed one of two
surveys: either the one allowing or the one precluding identification with the group. That is, they read a set of five
positive and five negative statements about their school and tallied their agreement with each set. Paired-samples t-
tests indicated that, when identification was impeded, participants agreed with fewer positive (M = 1.27, SD = 0.90)
than negative (M = 3.80, SD = 0.90) items, t (50) = -11.35, p < .001. When identification was enabled, participants
agreed with more positive (M = 3.94, SD = 1.30) than negative (M = 0.87, SD = 1.12) items, t (51) = -11.00, p < .
001. To form a measure of identification, we standardized and averaged the four items used in Study 2 with the
number of positive and negative (reversed) statements selected, α = .91. After imagining their reactions to a group of
people who identified highly with the school, participants reported less identification with the group when
identification was impeded (M = -0.46, SD = 0.65) relative to when it was enabled (M = 0.45, SD = 0.74), t (101) =
6.62, p < .001, d = 1.31, providing additional evidence that our manipulation effectively enabled and impeded
identification.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
44
the identification = enabled conditions, rank affected identification. However, in the
identification = impeded conditions, rank did not affect identification. The rank X identification
interaction was non-significant, F (1, 148) = 0.84, p = .36, ηp2 = .01.
Ethical views. We then examined how rank and identification affected views of how
ethical the group’s decision was. A marginally significant main effect emerged for rank, F (1,
148) = 3.15, p = .078, ηp2 = .02. A main effect emerged for identification as well, F (1, 148) =
4.21, p = .042, ηp2 = .03. Again, however, most important were the simple effects: When
identification was enabled, high-ranking participants (M = 3.87, SD = 1.20) saw the group’s
decision as significantly more ethical than low-ranking participants (M = 2.88, SD = 1.41), t (74)
= 3.27, p = .002, d = 0.76. When identification was impeded, no effect of high (M = 3.72, SD =
1.19) relative to low (M = 3.93, SD = 1.59) rank emerged, t (74) = -0.65, p = .52, d = 0.15. These
findings provided support for the idea that high-ranking individuals view the group’s decisions as
more ethical than do lower-ranking individuals, but only when identification with the group is
likely to emerge. The rank X identification interaction was also significant, F (1, 148) = 7.38, p =
.007, ηp2 = .05.
Principled dissent. To examine principled dissent, we used logistic regression, which is
appropriate for binary dependent measures. The main effects for rank (OR = 1.35, z = 0.36, p = .
55) and identification (OR = 2.11, z = 2.37, p = .12) were non-significant. Most relevant to our
hypotheses was the fact that holding higher rank reduced principled dissent, OR = 0.32, z = 4.69,
p = .03, when identification was enabled. When identification was impeded, no effect of rank
emerged, OR = 1.35, z = 0.36, p = .55. These findings were consistent with Hypothesis 2. An
interaction between rank and identification also emerged, OR = 0.24, z = 3.92, p = .048. Figure 3
displays the frequency of principled dissent by condition.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
45
We then examined strength of dissent. The main effects of rank (F [1, 148] = 0.85, p = .
36, ηp2 = .01) and identification (F [1, 148] = 0.52, p = .47, ηp2 = .004) were non-significant.
Similar to the above findings, when identification was enabled, high-ranking participants (M =
2.41, SD = 1.66) dissented less strongly than low-ranking participants (M = 3.34, SD = 2.00), t
(74) = -2.18, p = .03, d = 0.51. When identification was impeded, no difference in dissent
strength emerged between those of high (M = 2.86, SD = 1.78) and low (M = 2.47, SD = 1.73)
rank, t (74) = 0.96, p = .34, d = 0.22. A rank X identification interaction again emerged, F (1,
148) = 5.02, p = .03, ηp2 = .03.
Mediation analysis. Finally, we explored whether ethical views could help to explain
why high-ranking individuals engaged in less principled dissent than low-ranking individuals
when they were more likely to identify with the group, but not when they were less likely to do
so. When identification was enabled, the effect of high rank on principled dissent (p = .03)
dropped to non-significance (p = .92) when ethical views (OR = 0.20, z = 17.20, p < .001)
entered the logistic regression analysis. A bootstrapping analysis of mediation (Preacher &
Hayes, 2008) with 10,000 resamples produced a 95% confidence interval of (-3.37, -0.59),
providing evidence of significant mediation by ethical views. When identification was impeded,
the 95% confidence interval for ethical views included zero (-1.01, 2.29). These findings support
the notion that changes in ethical views help to explain why high- ranking individuals engage in
less principled dissent than low-ranking individuals when the effect of rank on identification is
likely to emerge.
Discussion
Study 3 examined our theoretical mechanism via moderation. That is, alongside
hierarchical rank, we manipulated conditions that would enable (or impede) identification with
HIERARCHICAL RANK AND PRINCIPLED DISSENT
46
the group to emerge. Consistent with Hypothesis 1, high-ranking individuals engaged in less
principled dissent when conditions enabled them to identify highly with the group that proposed
the unethical practice. No effects of hierarchical rank on principled dissent emerged when the
effect of rank on identification was impeded. Together, these findings support identification as
one reason that high-ranking leaders engage in less principled dissent than the people they are
responsible for overseeing.
Moreover, holding higher rank changed individuals’ attitudes toward the decisions within
the groups where they held higher rank. High-ranking individuals viewed the group’s decision as
more ethical than did lower-ranking individuals when identification with the group was enabled.
As a result, higher-ranking individuals may not desire to dissent on principled grounds,
compared to lower-ranking individuals.
General Discussion
Summary of Findings
Across three studies, holding a high-ranking position reduced principled dissent against
unethical practices. Increased identification partially explained the negative relation between
higher rank and principled dissent. That is, high-ranking individuals engaged in less principled
dissent in part because they identified with the group more than those with low-ranking.
Identification allowed high-ranking individuals to perceive the group’s practices as more ethical,
which hindered their ability to detect an ethical problem.
Strengths of the Research
This research has a number of strengths. First, its multi-method approach to examining
rank and principled dissent lends confidence that the relation emerges in the real world, as well
as within controlled settings. Second, it measures principled dissent using real behavior, not only
HIERARCHICAL RANK AND PRINCIPLED DISSENT
47
self-reports. Third, our findings explore principled dissent under two important conditions: when
high-ranking individuals are likely to identify with the group, and when they are not.
Contributions to Theories of Principled Dissent
We provide the first direct empirical test of an idea in Graham’s (1986) seminal theory
paper. She had predicted that being of higher level in the organization would increase (not
decrease) principled dissent. Our results diverge from her prediction due to a variable she saw as
central to principled dissent: awareness of an unethical practice. Holding higher rank prevented
people from perceiving an ethical problem, instead leading people to view the group’s decisions
as more ethical. Our findings highlight the importance of understanding how an unethical
practice is perceived.
To date, most research has studied principled dissent as whistleblowing (e.g., Miceli &
Near, 1984; Miceli, Near, & Dworkin, 1993). Graham (1986) recognized other forms of
principled dissent, such as actions taken to block unethical behavior and protest expressed to
others within the organization. Our results in Study 1 show that our effects emerge using a
reporting measure. Then, in Studies 2-4, we initiate attention to the latter form of principled
dissent. Protest or disagreement expressed to others may be a more common type of principled
dissent among high-ranking individuals, who have fewer people to report unethical activity to
and are responsible overseeing others’ decisions and leading committees that make collective
decisions. Expressing protest or disagreement could be an important form of principled dissent
under a variety of other conditions, such as when the ethical problem is ambiguous.
Conversations about the morality of a practice may be necessary to align perceptions because
individuals often overestimate the degree to which others share their ethical views (Flynn &
Wiltermuth, 2010).
HIERARCHICAL RANK AND PRINCIPLED DISSENT
48
By examining principled dissent, we contribute to a broader literature on corruption
controls (for a review, see Lange, 2008). Notably, most controls are implemented by high-
ranking individuals and directed at lower-ranking individuals. Our research suggests low-ranking
individuals may be better situated to perceive and act against unethical practices. Seminal
research attributed unethical practices to constraints acting on lower-ranking individuals, such as
immoral orders (Milgram, 1963; for a review, see Hamilton & Sanders, 1992) or organizational
processes (institutionalization, rationalization, and socialization) that prevent newcomers from
intervening (Ashforth & Anand, 2003). We document an additional explanation for unethical
practices. Unethical practices may persist in organizations because people charged with stopping
such practices identify too highly to carry out this responsibility.
Contributions to Theories of Rank and Power
Our research highlights the importance of attending to social identity theories (Ashforth
& Mael, 1989; Hogg & Terry, 2000; Pratt, 2001; Tajfel, 1982; Tajfel & Turner, 1985) when
studying rank and power in groups. Although early research on the Tennessee Valley Authority
predicted greater identification among individuals with higher rank in the organization (Brown,
1969), the idea has received only an indirect empirical test. Kreiner and Ashforth (2004) found a
positive relation between being a supervisor and identification, but the causal direction was
unclear. Moreover, Joshi and Fast (2013) focused on role identification, rather than group or
organizational identification. Our research is the first to establish a causal relation between
power and identification with a group or organization. 8 Understanding the effect of power on
group identification represents a significant theoretical advance in the study of power. Although
8 To better understand the different components of rank, we disentangled the effects of power and status on
principled dissent in Study 4 (see the Online Supplement). There, we found that principled dissent is lower for those
with elevated power, but not for those with elevated status, suggesting that power in particular discourages
principled dissent. Power, even without status, increased identification with the group.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
49
interest in hierarchy has recently surged (Anderson & Kennedy, 2012), we found no other studies
using group identification to predict behavior of powerful individuals.
Because power causes greater identification, blurring the boundaries between self and
group, high-ranking individuals are more inclined to accept the practices of the group that
accorded them power. Consequently, high-ranking individuals’ behavior is a function of not only
their own traits and goals, but also of the traits and goals of the groups they lead. For instance,
people who are by nature cooperative may become more competitive after attaining higher
power in a group with competitive norms, and vice versa.
Contributions to Theories of Identification and Ethical Behavior
High levels of identification have been theorized to lead to some negative outcomes, such
as reduced creativity, flexibility, and performance and increased conformity, dependency, and
vulnerability (Ashforth & Mael, 1989; Dutton & Dukerich, 1991; Michel & Jehn, 2003; Pratt,
2000). In addition, high identification (Umphress, Bingham, & Mitchell, 2010) has been
suspected to cause unethical behavior (Dukerich et al., 1998; Vadera & Pratt, 2013). Yet this idea
has received few empirical tests. The only empirical test we are aware of (Umphress et al., 2010)
used correlational data and self-reported measures of willingness to commit unethical behaviors.
We extend knowledge in two respects. First, our experimental designs are better able to
establish the causal role of identification. Second, we help to explain why identification leads to
unethical behavior. We find that changes in ethical views link identification to acceptance of
unethical behavior. In Study 3, identifying highly led high-ranking individuals to view an
unethical practice as more ethical than did others. Highly identified people may not be setting
aside their personal values to act unethically; instead, high identifiers see their group’s practices
as less unethical than do others. As a result, highly identified people may be susceptible to
HIERARCHICAL RANK AND PRINCIPLED DISSENT
50
“unintended unethicality,” (Tenbrunsel & Messick, 2004; Tenbrunsel & Smith-Crowe, 2008), or
inclined to believe they are acting ethically, even if they are not.
Limitations and Future Directions
The scope of high-ranking individuals’ conformity. Principled dissent is a form of non-
conformity (Warren, 2003), and one question then is whether high-ranking individuals conform
in other ways as well. For instance, they could engage in less dissent of other types by accepting
their groups’ ethical practices, agreeing with their groups’ strategic recommendations, or
rejecting creative ideas, meaning our theoretical perspective could apply more broadly.
Supporting this idea, Study 2 found holding higher rank to reduce unprincipled dissent (i.e.,
dissent against ethical practices) as well as principled dissent. Future research could examine
whether higher rank reduces of other types. Although this question is important, we focused on
principled dissent against unethical practices for a few reasons. First, the topic has applied
importance. Unethical practices have important consequences for society and are worthy of
exploration on their own accord. Second, unethical practices are often perceived as beneficial to
group outcomes. For instance, a Chief Financial Officer might over-state earnings in order to
protect her company’s share price, or the Head of Manufacturing may choose to ship product to a
customer earlier than requested in order to meet a sales goal, securing a holiday bonus for his
team. To the extent that unethical practices such as these are perceived to be helpful to the group,
the effects of higher-rank on dissent may be especially important in the domain of unethical
practices, relative to other domains (e.g., ethical practices, strategic decisions), because high-
identifiers are highly motivated to help their groups .
Exploring additional mechanisms. Although the current research focused on group
identification, other explanations may exist for why high-ranking individuals are less likely to
HIERARCHICAL RANK AND PRINCIPLED DISSENT
51
engage in principled dissent, as we reviewed earlier. Future research could examine the relative
importance of other factors relative to identification in reducing principled dissent. One
important possibility is that high-ranking individuals fear losing their privileged position,
whereas low-ranking individuals have less to lose. The prospect of losing a high status position is
very threatening (Marr & Thau, 2014; Pettit, Yong, & Spataro, 2010). If principled dissent is
perceived to put one’s status or legitimacy at risk, these concerns could help explain why high-
ranking individuals dissented less. Principled dissent could be perceived to violate the exchange
relation those in elevated positions have entered into by accepting status. Groups accord greater
status to those who contribute to collective goals, and higher status group members are expected
to continue contributing in accord with the group’s wishes (Willer, 2009). Our findings in Study
4, where principled dissent varied by level of power but not status, argue against this reasoning.
However, future research should explore this possibility in contexts where the possibility of
status loss is more vivid.
Additionally, rank and tenure are often highly correlated in organizations, and high-
ranking individuals have often made greater investments in the organization than low-ranking
individuals. Because of these investments, high-ranking individuals may feel irrational for
holding negative views of the organization, or its values, practices, or goals. Seeing the need for
principled dissent may then prove difficult for high-ranking individuals not only because they
identify more with the group, but because they feel a need to justify their past investments in the
organization (Staw, 1981). This need to justify past behavior may derive from a desire to avoid
cognitive dissonance (Festinger, 1957) caused by holding negative attitudes toward the
organization after investing in its success or by social pressures to hold thoughts consistent with
one’s past behavior (Staw, 1981). Cognitive dissonance is especially likely when high-ranking
HIERARCHICAL RANK AND PRINCIPLED DISSENT
52
individuals see the need to change policies, practices, or norms they themselves promoted
because they will feel highly responsible for any negative consequences of their past decisions
(Cooper, 1971). Thus, principled dissent may be embarrassing for high-ranking individuals.
Finally, future research could explore the relation between group and role identification.
Whereas we and Willer (2009) find that people identify more with a group when it affords higher
rank, prior research has found that people identify more with roles that afford greater power
(Joshi & Fast, 2013). Future research could clarify which form of identification precedes the
other and the consequences for behavior when the two forms of identification diverge.
Exploring additional moderators. We explored unethical practices that help the group (at
least in the short-term) because we believe this type unethical practice to be common in
organizations. However, unethical practices are sometimes harmful to the group (Vadera & Pratt,
2013). To the extent that an unethical practice is seen as harmful to a group, our proposed effects
may not hold. The normative conflict model has found that high-identifiers dissent more than
low-identifiers when they perceive a practice to be harmful to the group (Packer, 2009; Packer &
Chasteen, 2010; Packer & Miners, 2014). To the extent that an unethical practice is perceived to
harm the group, we would expect a reversal of the relationship between high-rank and principled
dissent, such that high-ranking individuals dissent more than low-ranking individuals against
unethical practices they perceive as harmful. When unethical practices are ambiguous with
respect to impact on the group, we would still expect our findings to hold in light of the way
identification changed ethical views in Study 3. Supporting this idea, Study 1 examined illegal
and wasteful activity, and found less principled dissent among high-ranking individuals. Illegal
and wasteful practices are likely to be harmful to the group in the long-term, but identification
may have shaped perceptions of whether the practices were harmful, much like it shaped ethical
HIERARCHICAL RANK AND PRINCIPLED DISSENT
53
views in Study 3. Future research could also explore situations where holding higher-rank does
not blind individuals to ethical problems. In particular, interventions that generate higher-level
construal could help to mitigate this tendency (Packer et al., 2014).
Whether unethical acts are prevalent practices, or one-off misconduct could be another
important moderator. We examined principled dissent in the context of practices, and we expect
our effects to be stronger for practices than for one-off misconduct. Identification may not blind
high-ranking individuals to the group’s problems when misconduct is one-off, rather than widely
accepted.
Gender effects. Prior research has found evidence of ethical differences between women
and men (e.g., Dreber & Johannesson, 2008; Gilligan, 1982; Franke et al., 1997; Haselhuhn &
Wong, 2012; Jaffee & Hyde, 2000; Kennedy & Kray, 2013; Kray & Haselhuhn, 2012; Robinson,
Lewicki, & Donahue, 2000). Our research only partially replicated these findings. Women did
object to ethical practices less often than men, consistent with prior research. However, women
did not object to unethical practices more often than men. Future research could explore why the
gender effects diverged for principled versus unprincipled dissent. It could be that principled
dissent is highly assertive and women fear backlash for engaging in it (cf. Rudman & Phelan,
2008), or women might fear losing their high-ranking positions to a greater extent than do men.
Conclusion
In our research, those who held positions of higher rank were less likely to engage in
principled dissent against unethical practices. Higher-ranking individuals’ greater identification
with the group partially explained why. Consistent with prior theorizing (Ashforth & Anand,
2003; Mayer et al., 2013; Smith-Crowe & Warren, 2014; Tenbrunsel & Messick, 2004;
Tenbrunsel & Smith-Crowe, 2008; Trevino, 1986; Trevino et al., 2008; Vadera & Pratt, 2013;
HIERARCHICAL RANK AND PRINCIPLED DISSENT
54
Warren & Smith-Crowe, 2008), our findings suggest that unethical behavior in organizations
may have more complex origins than simply self-interest or negligent oversight. Individuals at
the top of the organizational hierarchy may be blinded by identification, constraining their ability
to discern and stop unethical behavior. Because high-ranking individuals view the group’s
practices as more ethical, the common approach of allocating them responsibility for ensuring
ethical behavior in the organization may be misguided. Lower-ranking individuals who also
identify highly with the group or organization could provide better ethical guidance.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
55
References
Anderson, C., & Berdahl, J. L. (2002).The experience of power: Examining the effects of power
on approach and inhibition tendencies. Journal of Personality and Social Psychology, 83,
1362-1377.
Anderson, C., & Galinsky, A. D. (2006). Power, optimism, and risk-taking. European Journal of
Social Psychology, 36, 511-536.
Anderson, C., Hildreth, J. A. D. & Howland, L. (2015). Is the desire for status a fundamental
human motive? A review of the empirical literature. Psychological Bulletin, 141, 574-
601.
Anderson, C., & Kennedy, J. A. (2012). Micropolitics: A new model of status hierarchies in
teams. In M. A. Neale, & E. A. Mannix (Eds.), Research on managing groups and teams
(vol. 15, pp. 49-80). Bingley, UK: Emerald.
Anicich, E., M., Fast, N. J., Halevy, N., & Galinsky, A. D. (2015). When the bases of social
hierarchy collide: Power without status drives interpersonal conflict. Organization
Science, 27, 123-140.
Antilla, S. (2002). Tales from the boom-boom room. New York: Bloomberg.
Asch, S. E. (1956). Studies of independence and conformity: A minority of one against a
unanimous majority. Psychological Monographs, 70, 1-71.
Ashforth, B. E., & Anand, V. (2003). The normalization of corruption in organizations. In B. M.
Staw, & R. M. Kramer (Eds.), Research in Organizational Behavior (vol. 25, pp. 1-52).
Oxford, UK: Elsevier.
Ashforth, B. E., & Mael, F. (1989). Social identity theory and the organization. Academy of
Management Review, 14, 20-39.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
56
Bales, R. F., Strodtbeck, F. L., Mills, T. M., & Roseborough, M. E. (1951). Channels of
communication in small groups. American Sociological Review, 16, 461-468.
Baumhart, R. C. (1961). How ethical are businessmen? Harvard Business Review, 39, 6-19.
Becker, H. S. (1963). Outsiders: Studies in the sociology of deviance. New York: Free Press.
Berger, J., Cohen, B. P., & Zelditch, M. (1972). Status characteristics and social interaction.
American Sociological Review, 37, 241-255.
Berger, J., Webster, M., Ridgeway, C., & Rosenholtz, S. J. (1986). Status cues, expectations, and
behaviors. In E.J. Lawler (Ed.), Advances in group processes (vol. 3, pp. 1–22).
Greenwich, CT: JAI Press.
Binning, K. R., Brick, C., Cohen, G. L., & Sherman, D. K. (2015). Going along versus getting it
right: The role of self-integrity in political conformity. Journal of Experimental Social
Psychology, 56, 73-88.
Blader, S. L., & Chen, Y.-R. (2011). What influences how higher-status people respond to lower-
status others? Effects of procedural fairness, outcome favorability, and concern about
status. Organization Science, 22, 1040-1060.
Blader, S. L., & Chen, Y.-R. (2012). Differentiating the effects of status and power: A justice
perspective. Journal of Personality and Social Psychology, 102, 994-1014.
Boivie, S., Lange, D., McDonald, M. L., & Westphal, J. D. (2011). Me or I: The effects of CEO
organizational identification on agency costs. Academy of Management Journal, 54, 551-
576.
Bowles, H. R., & Gelfand, M. (2010). Status and the evaluation of workplace deviance.
Psychological Science, 21, 49-54.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
57
Brenner, S. N., & Molander, E. A. (1977). Is the ethics of business changing? Harvard Business
Review, 55, 55-71.
Brief, A. P., Buttram, R. T., & Dukerich, J. M. (2001). Collective corruption in the corporate
world: Toward a process model. In M. E. Turner (Ed.), Groups at Work: Theory and
Research (pp. 471–499). Mahwah, NJ: Erlbaum.
Brown, M. E. (1969). Identification and some conditions of organizational involvement.
Administrative Science Quarterly, 14, 346-355.
Carmeli, A., Shalom, R., Weisberg, J. (2007). Considerations in organizational career
advancement: What really matters. Personnel Review, 36, 190-205.
Carney, D. R., Cuddy, A. J. C., & Yap. A. J. (2010). Power posing: Brief nonverbal displays
affect neuroendocrine levels and risk tolerance. Psychological Science, 21, 1363-1368.
Carney, D. R., Dubois, D., Nichiporuk, N., ten Brinke, L., Rucker, D. D., & Galinsky, A. D.
(2015). The deception equilibrium: The powerful are better liars but the powerless are
better lie detectors. Manuscript submitted for publication.
Catalyst (2013). Catalyst Quick Take: Statistical overview of women in the workplace.
http://www.catalyst.org/knowledge/statistical-overview-women-workplace
Cavanaugh, G. F., Moberg, D. J., & Velasquez, M. (1981). The ethics of organizational politics.
Academy of Management Review, 6, 363-374.
Chatman, J.A. (1991). Matching people and organizations: Selection and socialization in public
accounting firms. Administrative Science Quarterly, 36, 459-484.
Chatman, J. A., & Flynn, F. J. (2005). Full-cycle micro-organizational behavior research.
Organization Science, 16, 434-447.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
58
Chen, S., Lee-Chai, A. Y., & Bargh, J. A. (2001). Relationship orientation as a moderator of the
effects of social power. Journal of Personality & Social Psychology, 80, 173-187.
Cohen, T. R., Gunia, B. C., Kim-Jun, S. Y., & Murnighan, J.K. (2009). Do groups lie more than
individuals? Honesty and deception as a function of strategic self-interest. Journal of
Experimental Social Psychology, 45, 1321-1324.
Cooper, J. (1971). Personal responsibility and dissonance: The role of foreseen consequences: An
integration and extension of the "forced compliance" literature. Journal of Personality
and Social Psychology, 18, 354-363.
DeCelles, K. A., DeRue, D. S., Margolis, J. D., & Ceranic, T. L. (2012). Does power corrupt or
enable? When and why power facilitates self-interested behavior. Journal of Applied
Psychology, 97, 681-689.
Detert, J. R., Burris, E. R., Harrison, D. A., & Martin, S. R. (2013). Voice flows to and around
leaders: Understanding when units are helped or hurt by employee voice. Administrative
Science Quarterly, 58, 624-668.
DeWall, C. N., Baumeister, R. F., Mead, N. L., & Vohs, K. D. (2011). How leaders self-regulate
their task performance: Evidence that power promotes diligence, depletion, and disdain.
Journal of Personality and Social Psychology, 100, 47-65.
Dreber, A., & Johannesson, M. (2008). Gender differences in deception. Economics Letters, 99,
197-199.
Dukerich, J. M., Golden, B. R., & Shortell, S. M. (2002). Beauty is in the eye of the beholder:
The impact of organizational identification, identity, and image on the cooperative
behaviors of physicians. Administrative Science Quarterly, 47, 507-533.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
59
Dukerich, J. M., Kramer, R., & Parks, J. M. (1998). The dark side of organizational
identification. In D. A. Whetten, & P. C. Godfrey (Eds.), Identity in organizations:
Building theory through conversations (pp. 245-256). Thousand Oaks, CA: Sage.
Dutton, J. E., & Dukerich, J. M. (1991). Keeping an eye on the mirror: Image and identity in
organizational adaptation. Academy of Management Journal, 34, 517-554.
Dutton, J. E., Dukerich, J. M., & Harquail, C. V. (1994). Organizational images and member
identification. Administrative Science Quarterly, 39, 239-263.
Ellemers, N., Spears, R., & Doosje, B. (1997). Sticking together or falling apart: In-group
identification as a psychological determinant of group commitment versus individual
mobility. Journal of Personality and Social Psychology, 72, 617-626.
Emerson, R. M. (1962). Power-dependence relations. American Sociological Review, 27, 31-41.
Etzioni, A. (1961). A comparative evaluation of complex organizations: On power, involvement
and their correlates. New York: Free Press.
Farrell, D., & Peterson, J. C. (1982). Patterns of political behavior in organizations. Academy of
Management Review, 7, 403-412.
Fast, N., Gruenfeld, D. H., Sivanathan, N., and Galinsky, A. D. (2009). Illusory control: A
generative force behind power's far-reaching effects. Psychological Science, 20, 502-508.
Fast, N. J., Halevy, N., & Galinsky, A. D. (2012). The destructive nature of power without status.
Journal of Experimental Social Psychology, 48, 391-394.
Festinger. L. (1950). Informal social communication. Psychological Review, 57, 271-282.
Festinger, L. (1957). A theory of cognitive dissonance. Palo Alto, CA: Stanford University Press.
Fiske, S. T., & Berdahl, J. (2007). Social power. In A. Kruglanski & E. T. Higgins (Eds.), Social
psychology: Handbook of basic principles (2nd ed.): 678-692. New York: Guilford Press.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
60
Flynn, F. J., Reagans, R. E., & Amanatullah, E. T., & Ames, D. R. (2006). Helping one’s way to
the top: Self-monitors achieve status by helping others and knowing who helps whom.
Journal of Personality and Social Psychology, 91, 1123-1137.
Flynn, F. J., & Wiltermuth, S. S. (2010). Who’s with me? False consensus, brokerage, and ethical
decision making in organizations. Academy of Management Journal, 53, 1074-1089.
Fragale, A. R., Overbeck, J. R., & Neale, M. A. (2011). Resources versus respect: Social
judgments based on targets’ power and status positions. Journal of Experimental Social
Psychology, 47, 767-775.
Fragale, A. R., Sumanth, J. J., Tiedens, L. Z., & Northcraft, G. B. (2012). Appeasing equals:
Lateral deference in organizational communication. Administrative Science Quarterly,
57, 373-406.
Franke, G. R., Crown, D. F., & Spake, D. F. (1997). Gender differences in ethical perceptions of
business practices: A social role theory perspective. Journal of Applied Psychology, 82,
920-934.
Galinsky, A. D., Gruenfeld, D. H., & Magee, J. C. (2003). From power to action. Journal of
Personality and Social Psychology, 85, 453-466.
Galinsky, A. D., Magee, J. C., Inesi, M. E., & Gruenfeld, D. H. (2006). Power and perspectives
not taken. Psychological Science, 17, 1068-1074.
Galinsky, A. D., Magee, J. C., Gruenfeld, D. H., Whitson, J. A., & Liljenquist, K. A. (2008).
Power reduces the press of the situation: Implications for creativity, conformity, and
dissonance. Journal of Personality and Social Psychology, 95, 1450-1466.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
61
Georgesen, J. C., & Harris, M. J. (1998). Why’s my boss always holding me down? A meta-
analysis of power effects on performance evaluations. Personality and Social Psychology
Review, 2, 184-195.
Gilligan, C. (1982). In a different voice: Psychological theory and women’s development.
Cambridge, MA: Harvard University Press.
Gneezy, U. (2005). Deception: The role of consequences. American Economic Review, 95, 384-
394.
Goldstein, N. J., & Hays, N. A. (2011). Illusory power transference: The vicarious experience of
power. Administrative Science Quarterly, 56, 593-621.
Graham, J. W. (1986). Principled organizational dissent: A theoretical essay. In B. M. Staw, & L.
L. Cummings (Eds.), Research in Organizational Behavior (vol. 8, pp. 1-52). Greenwich,
CT: JAI Press.
Greenaway, K. H., Haslam, S. A., Cruwys, T., Branscombe, N. R., Ysseldyk, R., & Heldreth, C.
(2015). From “we” to “me”: Group identification enhances perceived personal control
with consequences for health and well-being. Journal of Personality and Social
Psychology, 109, 53-74.
Gruenfeld, D. H., Inesi, M. E., Magee, J. C., & Galinsky, A. D. (2008). Power and the
objectification of social targets. Journal of Personality and Social Psychology, 95, 111-
127.
Guimond, S. (1995). Encounter and metamorphosis: The impact of military socialisation on
professional values. Applied Psychology: An International Review, 44, 251-275.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
62
Guimond, S., Dambrun, M., Michinov, N., & Duarte, S. (2003). Does social dominance generate
prejudice? Integrating individual and contextual determinants of intergroup cognitions.
Journal of Personality and Social Psychology, 84, 697-721.
Hamermesh, R., Whittemore, A., & Sherman, E. (2010). The Jenner situation. Harvard Business
School case study. Boston: President and Fellows of Harvard College.
Hamilton, V. L., & Sanders, J. (1992). Responsibility and risk in organizational crimes of
obedience. In B. M. Staw, & L. L. Cummings (Eds.), Research in Organizational
Behavior (vol. 14, pp. 49-90). Greenwich, CT: JAI Press.
Hamilton, V. L., & Sanders, J. (1995). Crimes of obedience and conformity in the workplace:
Surveys of Americans, Russians, and Japanese. Journal of Social Issues, 51, 67-88.
Haselhuhn, M. P., & Wong, E. M. (2012). Bad to the bone: Facial structure predicts unethical
behaviour. Proceedings of the Royal Society B, 279, 571-576.
Hayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable mediation,
moderation, and conditional process modeling [White paper]. Retrieved from
http://www.afhayes.com/public/process2012.pdf
Heider, F. (1958). The psychology of interpersonal relations. Wiley: New York.
Hertel, G., & Kerr, N. L. (2001). Priming in-group favoritism: The impact of normative scripts in
the minimal group paradigm. Journal of Experimental Social Psychology, 37, 316-324.
Hirschman, A. O. (1970). Exit, voice, and loyalty. Cambridge, MA: Harvard University Press.
Hogg, M. A., Fielding, K. S., & Darley, J. (2005). Fringe dwellers: Processes of deviance and
marginalization in groups. In D. Abrams, M. A. Hogg, & J. M. Marques (Eds.), The
social psychology of inclusion and exclusion (pp. 191-210). New York: Psychology Press.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
63
Hogg, M. A., & Terry, D. J. (2000). Social identity and self-categorization processes in
organizational contexts. Academy of Management Review, 25, 121-140.
Hollander, E. P. (1958). Conformity, status, and idiosyncrasy credit. Psychological Review, 65,
117-127.
Hollander, E. P. (1961). Some effects of perceived status on responses to innovative behavior.
Journal of Abnormal and Social Psychology, 63, 247-250.
Jaffee, S., & Hyde, J. S. (2000). Gender differences in moral orientation: A meta-analysis.
Psychological Bulletin, 126, 703-726.
Janis, I. L. (1972). Victims of groupthink: A psychological study of foreign-policy decisions and
fiascoes. Oxford, United Kingdom: Houghton Mifflin.
Jetten, J., Postmes, T., & McAuliffe, B. J. (2002). ‘We're all individuals’: Group norms of
individualism and collectivism, levels of identification and identity threat. European
Journal of Social Psychology, 32, 189-207.
Joshi, P. D., & Fast, N. J. (2013). I am my (high-ranking) role: Power and role identification.
Personality and Social Psychology Bulletin, 39, 898-910.
Kahneman, D., Knetsch, J. L., & Thaler, R. (1986). Fairness as a constraint on profit-seeking:
Entitlements in the market. American Economic Review, 76, 728-741.
Karp, R., Gold, D., & Tan, M. (1998). Bullard houses In J. M. Brett (Ed.), Teaching materials for
negotiations and decision making: 12. Evanston, IL: Northwestern University, Dispute
Resolution Research Center.
Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and inhibition.
Psychological Review, 110, 265-284.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
64
Kennedy, J. A., & Kray, L. J. (2013). Who is willing to sacrifice ethical values for money and
social status?: Gender differences in reactions to ethical compromises. Social
Psychological and Personality Science, 5, 52-59.
Kifer, Y., Heller, D., Perunovic, W. Q. E., & Galinsky, A. D. (2013). The good life of the
powerful: The experience of power and authenticity enhances subjective well-being.
Psychological Science, 24, 280-288.
Kipnis, D. (1972). Does power corrupt? Journal of Personality and Social Psychology, 24, 33-
41.
Kouchaki, M., Smith-Crowe, K., Brief, A. P., & Sousa, C. (2013). Seeing green: Mere exposure
to money triggers a business decision frame and unethical outcomes. Organizational
Behavior and Human Decision Processes, 121, 53-61.
Kray, L. J., & Haselhuhn, M. P. (2012). Male pragmatism in negotiators’ ethical reasoning.
Journal of Experimental Social Psychology, 48, 1124-1131.
Kreiner, G. E., & Ashforth, B. E. (2004). Evidence toward an expanded model of organizational
identification. Journal of Organizational Behavior, 25, 1-27.
Lammers, J., Stapel, D. A., & Galinsky, A. D. Power increases hypocrisy: Moralizing in
reasoning, immorality in behavior. Psychological Science, 21, 737-744.
Lange, D. (2008). A multidimensional conceptualization of organizational corruption control.
Academy of Management Review, 33, 710-729.
Leach, C. W., Ellemers, N., & Barreto, M. (2007). Group virtue: The importance of morality (vs.
competence and sociability) in the positive evaluation of in-groups. Journal of
Personality and Social Psychology, 93, 234-249.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
65
Leary, M. R., Cottrell, C. A., & Phillips, M. (2001). Deconfounding the effects of dominance and
social acceptance on self-esteem. Journal of Personality and Social Psychology, 81, 898-
909.
Mackie, D., & Cooper, J. (1984). Attitude polarization: Effects of group membership. Journal of
Personality and Social Psychology, 46, 575-585.
Mael, F., & Ashforth, B. E. (1992). Alumni and their alma mater: A partial test of the
reformulated model of organizational identification. Journal of Organizational Behavior,
13, 103-123.
Magee, J. C., & Galinsky, A. D. (2008). Social hierarchy: The self-reinforcing nature of power
and status. In J. P. Walsh, & A. P. Brief (Eds.), Academy of Management Annals (vol. 2,
pp. 351-398). New York: Routledge.
Magee, J. C., Kilduff, G. J., & Heath, C. (2011). On the folly of principal's power: Managerial
psychology as a cause of bad incentives. In A. P. Brief, & B. M. Staw (Eds.), Research in
Organizational Behavior (vol. 31, pp. 25-41). New York: Elsevier.
Marmot, M. (2004). The status syndrome: How social standing affects our health and longevity.
New York: Owl Books.
Marr, J. C., & Thau, S. (2014). Falling from great (and not-so-great) heights: How initial status
position influences performance after status loss. Academy of Management Journal, 57,
223-248.
Mast, M. S., Jonas, K., & Hall, J. A. (2009). Give a person power and he or she will show
interpersonal sensitivity: The phenomenon and its why and when. Journal of Personality
and Social Psychology, 97, 835-850.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
66
Mayer, D. M., Kuenzi, M., Greenbaum, R., Bardes, M., & Salvador, R. (2009). How low does
ethical leadership flow? Test of a trickle-down model. Organizational Behavior and
Human Decision Processes, 108, 1-13.
Mayer, D. M., Nurmohamed, S., Trevino, L. K., Shapiro, D. L., & Schminke, M. (2013).
Encouraging employees to report unethical conduct internally: It takes a village.
Organizational Behavior and Human Decision Processes, 121, 89-103.
McCullagh, D. (2006). FAQ: Behind the stock options uproar. CNet. http://news.cnet.com/FAQ-
Behind-the-stock-options-uproar/2100-1014_3-6098457.html
McDonald, M. L., & Westphal, J. D. (2011). Getting by with the advice of their friends: CEOs’
advice networks and firms’ strategic responses to poor performance. Administrative
Science Quarterly, 48, 1-32.
Miceli, M. P., & Near, J. P. (1984). The relationship among beliefs, organizational position, and
whistle-blowing status: A discriminant analysis. Academy of Management Journal, 27:
687-705.
Miceli, M. P., & Near, J. P. (1992). Blowing the whistle: The organizational and legal
implications for companies and their employees. New York: Lexington.
Miceli, M. P., & Near, J. P. (2002). What makes whistle-blowers effective? Three field studies.
Human Relations, 55, 455-479.
Miceli, M. P., Near, J. P., & Dworkin, T. M. (1993). Explaining the whistle-blowing process:
Suggestions from power theory and justice theory. Organization Science, 4, 393-411.
Michel, A. (2011). Transcending socialization: A nine-year ethnography of the body’s role in
organizational control and knowledge workers’ transformation. Administrative Science
Quarterly, 56, 325-368.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
67
Michel, A. A., & Jehn, K. E. (2003). The dark side of identification: Overcoming identification-
induced performance impediments. Research on Managing Groups and Teams, 5, 189-
219.
Milgram, S. (1963). Behavioral study of obedience. Journal of Abnormal and Social Psychology,
67, 371-378.
Moore, C., & Gino, F. (2013). Ethically adrift: How others pull our moral compass from true
North and how we can fix it. In A. P. Brief & B. M. Staw (Eds.), Research in
Organizational Behavior (vol. 33, pp. 53-77). San Diego, CA: Elsevier.
Motavalli, J. (2010). At the Toyota hearing: Remembering the Pinto. New York Times. Retrieved
from http://wheels.blogs.nytimes.com/2010/02/25/at-the-toyota-hearing-remembering-
the-pinto/
Near, J. P., & Miceli, M. P. (1985). Organizational dissidence: The case of whistle-blowing.
Journal of Business Ethics, 4, 1-16.
Near, J. P., & Miceli, M. P. (1995). Effective whistle-blowing. Academy of Management Review,
20, 679-708.
Nemeth, C. J., & Staw, B. M. (1989). The tradeoffs of social control and innovation in groups
and organizations. In L. Berkowitz (Ed.), Advances in experimental social psychology,
(vol. 22, pp. 175-209). San Diego, CA: Academic Press.
O’Fallon, M. J., & Butterfield, K. D. (2005). A review of the empirical ethical decision-making
literature: 1996-2003. Journal of Business Ethics, 59, 375-413.
O’Reilly, C. A. III., & Chatman, J. A. (1986). Organizational commitment and psychological
attachment: The effects of compliance, identification, and internalization of prosocial
behavior. Journal of Applied Psychology, 71, 492-499.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
68
Packer, D. J. (2008). On being both with us and against us: A normative conflict model of dissent
in social groups. Personality and Social Psychology Review, 12, 50-72.
Packer, D. J. (2009). Avoiding groupthink: Whereas weakly identified members remain silent,
strongly identified members dissent about collective problems. Psychological Science,
20, 546-548.
Packer, D. J., & Chasteen, A. L. (2010). Loyal deviance: Testing the normative conflict model of
dissent in social groups. Personality and Social Psychology Bulletin, 36, 5-18.
Packer, D. J., Fujita, K., & Chasteen, A. L. (2014). The motivational dynamics of dissent
decisions: A goal-conflict approach. Social Psychological and Personality Science, 5, 27-
34.
Packer. D. J., Fujita, K., & Herman, S. (2013). Rebels with a cause: A goal conflict approach to
understanding when conscientious people dissent. Journal of Experimental Social
Psychology, 49, 927-932.
Packer, D. J., & Miners, C. T. H. (2014). Tough love: The normative conflict model and a goal
system approach to dissent decisions. Social and Personality Psychology Compass, 8,
354-373.
Patsuris, P. (2002). The corporate scandal sheet. Forbes.com.
http://www.forbes.com/2002/07/25/accountingtracker.html
Pettit, N. C., Yong, K., & Spataro, S. E. (2010). Holding your place: Reactions to the prospect of
status gains and losses. Journal of Experimental Social Psychology, 46, 396-401.
Pillutla, M. M., & Chen, X. P. (1999). Social norms and cooperation in social dilemmas: The
effects of context and feedback. Organizational Behavior and Human Decision
Processes, 78, 81-103.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
69
Pitesa, M., & Thau, S. (2013). Compliant sinners, obstinate saints: How power and self-focus
determine the effectiveness of social influences in ethical decision making. Academy of
Management Journal, 56, 635-658.
Pratt, M. G. (2000). The good, the bad, and the ambivalent: Managing identification among
Amway distributors. Administrative Science Quarterly, 45, 456–493.
Pratt, M. G. (2001). Social identity dynamics in modern organizations: An organizational
psychology/ organizational behavior perspective. In M. Hogg & D. J. Terry (Eds.), Social
identity processes in organizational contexts (pp. 13-30). Philadelphia: Psychology Press.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and
comparing indirect effects in multiple mediator models. Behavior Research Methods, 40,
879-891.
Reicher, S. D., Haslam, S. A., & Smith, J. R. (2012). Working toward the experimenter:
Reconceptualizing disobedience within the Milgram paradigm as identification-based
followership. Perspectives on Psychological Science, 7, 315-324.
Robinson, R. J., Lewicki, R. J., & Donahue, E. M. (2000). Extending and testing a five factor
model of ethical and unethical bargaining tactics: Introducing the SINS scale. Journal of
Organizational Behavior, 21, 649-664.
Rudman, L. A., & Phelan, J. E. (2008). Backlash effects for disconfirming gender stereotypes in
organizations. In A. P. Brief, & B. M. Staw (Eds.), Research in organizational behavior
(28, pp. 61–79). New York, NY: Elsevier.
Schweitzer, M. E., Ordonez, L., & Douma, B. (2004). Goal setting as a motivator of unethical
behavior. Academy of Management Journal, 47, 422-432.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
70
Sechrist, G. B., & Young, A. F. (2011). The influence of social consensus information on
intergroup attitudes: The moderating effects of ingroup identification. Journal of Social
Psychology, 151, 674-695.
Sherman, G. D., Lee, J. J., Cuddy, A. J. C., Renshon, J., Oveis, C., Gross, J. J., & Lerner, J. S.
(2012). Leadership is associated with lower levels of stress. Proceedings of the National
Academy of Sciences, 109, 17903-17907.
See, K. E., Morrison, E. W., Rothman, N. B., & Soll, J. B. (2011). The detrimental effects of
power on confidence, advice taking, and accuracy. Organizational Behavior and Human
Decision Processes, 116, 272-285.
Sims, R. R. (1992). The challenge of unethical behavior in organizations. Journal of Business
Ethics, 11, 505-513.
Sims, R. R., & Brinkman. J. (2002). Leaders as moral role models: The case of John Gutfreund
at Salomon Brothers. Journal of Business Ethics, 35, 327-339.
Smith-Crowe, K., & Warren, D. E. (2014). The emotion-evoked collective corruption model: The
role of emotion in the spread of corruption within organizations. Organization Science,
25, 1154-1171.
Staw, B. M. (1981). The escalation of commitment to a course of action. Academy of
Management Review, 6, 577-587.
Staw, B. M., & Boetteger, R. D. (1990). Task revision: A neglected form of work performance.
Academy of Management Journal, 33, 534-559.
Stouffer, S. A., Suchman, E. A., DeVinney, L. C., Star, S. A., & Williams, R. M., Jr. (1949). The
American soldier: Adjustment during wartime life (vol. 1). Princeton, NJ: Princeton
University Press.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
71
Strudler, A., & Warren, D. E. (2001). Authority, heuristics, and the structure of excuses. In J. M.
Darley, D. M. Messick, & T. Tyler (Eds.), Social influence and ethics (pp. 155-173).
Mahwah, NJ: Lawrence Erlbaum.
Tajfel, H. (1982). Social identity and intergroup relations. Cambridge, United Kingdom:
Cambridge University Press.
Tajfel, H., & Turner, J. C. (1985). The social identity theory of intergroup behavior. In S.
Worchel & W. G. Austin (Eds.), Psychology of intergroup behavior (pp. 7-24). Chicago:
Nelson-Hall.
Tannenbaum, A. S., Kavcic, B., Rosner, M., Vianello, M., & Wieser, G. (1974). Hierarchy in
organizations. San Francisco: Jossey-Bass.
Tay, L., & Diener, E. (2011). Needs and subjective well-being around the world. Journal of
Personality and Social Psychology, 101, 354-365.
Tenbrunsel, A. E., & Messick, D. M. (2004). Ethical fading: The role of self-deception in
unethical behavior. Social Justice Research, 17, 223-236.
Tenbrunsel, A. E., & Smith-Crowe, K. (2008). Ethical decision-making: Where we’ve been and
where we’re going. In J. P. Walsh, & A. P. Brief (Eds.), Academy of Management Annals
(vol. 2, pp. 545-607). New York: Routledge.
Tost, L. P. (2015). When, why, and how do powerholders ‘‘feel the power’’? Examining the links
between structural and psychological power and reviving the connection between power
and responsibility. In A. P. Brief & B. M. Staw (Eds.), Research in Organizational
Behavior (vol. 35, pp. 29-56).
HIERARCHICAL RANK AND PRINCIPLED DISSENT
72
Tost, L. P., Gino, F., & Larrick, R. P. (2013). When power makes others speechless: The negative
impact of leader power on team performance. Academy of Management Journal, 56,
1465-1486.
Trevino, L. K. (1986). Ethical decision making in organizations: A person-situation interactionist
model. Academy of Management Review, 11, 601-617.
Trevino, L. K., Weaver, G. R., & Brown, M. E. (2008). It’s lovely at the top: Hierarchical levels,
identities, and perceptions of organizational ethics. Business Ethics Quarterly, 18, 233-
252.
Tsui, L. (1999). Courses and instruction affecting critical thinking. Research in Higher
Education, 40, 185-200.
Tyler, T. R., & Blader, S. L. (2003). The group engagement model: Procedural justice, social
identity, cooperative behavior. Personality and Social Psychology Review, 7, 349-361.
Umphress, E. E., Bingham, J. B., & Mitchell, M. S. (2010). Unethical behavior in the name of
the company: The moderating effect of organizational identification and positive
reciprocity beliefs on unethical pro-organizational behavior. Journal of Applied
Psychology, 95, 769-780.
Vadera, A.K. & Pratt, M.G. (2013). Love, hate, ambivalence, or indifference?: A conceptual
examination of workplace crime and organizational identification. Organization Science,
24, 172-188.
Van Kleef, G. A., Oveis, C., van der Lowe, I., LuoKogan, A., Goetz, J., & Keltner, D. (2008).
Power, distress, and compassion: Turning a blind eye to the suffering of others.
Psychological Science, 19, 1315-1322.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
73
Walumbwa, F. O., Mayer, D. M., Wang, P., Wang, H., Workman, K., & Christensen, A. L. (2011).
Linking ethical leadership to employee performance: The roles of leader–member
exchange, self-efficacy, and organizational identification. Organizational Behavior and
Human Decision Processes, 115, 204-213.
Warren, D. E. (2003). Constructive and destructive deviance in organizations. Academy of
Management Review, 28, 622-632.
Warren, D. E., & Smith-Crowe, K. (2008). Deciding what’s right: The role of external sanctions
and embarrassment in shaping moral judgments in the workplace. In A. P. Brief, & B. M.
Staw (Eds.), Research in organizational behavior (vol. 28, pp. 81-105). San Diego, CA:
Elsevier.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures
of positive and negative affect: The PANAS scales. Journal of Personality and Social
Psychology, 54, 1063-1070.
Weber, M. (1948). Essays in sociology. H. H. Gerth and C. W. Mills (Eds.). Oxford, United
Kingdom: Routledge.
Willer, R. (2009). Groups reward individual sacrifice: The status solution to the collective action
problem. American Sociological Review, 74, 23-43.
Wiltermuth, S., & Flynn, F. (2013). Power, moral clarity, and punishment in the workplace.
Academy of Management Journal, 56, 1002-1023.
Wojciszke, B., & Struzynska-Kujalowicz, A. (2007). Power influences self-esteem. Social
Cognition, 25, 472-494.
Zimmerer, T. W., & Preston, P. L. (1976). Plasma International. In R. D. Hay, E. R. Gray, & J. E.
Gates (Eds.), Business and society: Cases and text. Cincinnati, OH: South-Western.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
74
Zou, X., Morris, M. W., & Benet-Martinez, V. (2008). Identity motives and cultural priming:
Cultural (dis)identification in assimilative and contrastive responses. Journal of
Experimental Social Psychology, 44, 1151-1159.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
75
TABLE 1
Correlations among Variables in Study 1
Variable M SD 1234567
1. Hierarchical rank 3.04 0.71 --
2. Female gender 0.41 0.49 .03** --
3. Ethnicitya-- -- -.09*** .14*** --
4. Education 4.48 1.68 .41*** -.24*** -.13*** --
5. Tenure 4.41 1.82 .15*** -.15*** -.05*** -.05*** --
6. Age 4.11 1.33 .09*** -.14*** -.07*** -.02* .56*** --
7. Knowledge of protections 2.08 0.96 .12*** -.09*** -.00 .05*** .18*** .14*** --
8. Principled dissent 0.16 0.37 -.04*** -.03** -.05*** .04*** -.02* -.04*** .02*
* p < .05. ** p < .01. *** p < .001.
a Dummy variable coded 0 for Caucasians and 1 for other ethnicities.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
76
HIERARCHICAL RANK AND PRINCIPLED DISSENT
77
TABLE 2
Logistic Regression Models Predicting Reporting Unethical Behavior in Study 1a
Variable Model 1 Model 2
Controls
Female gender 0.87* (.05) 0.97 (.06)
Ethnicityb0.76*** (.05) 0.75*** (.05)
Education 1.04* (.02) 1.09*** (.02)
Tenure 1.00 (.02) 1.01 (.02)
Age 0.90*** (.02) 0.91*** (.02)
Knowledge of protections 1.07* (.03) 1.08** (.03)
Hierarchical Rank
Level 2 0.76† (.11)
Level 3 0.52*** (.06)
Level 4 0.48*** (.07)
Level 5 0.36*** (.08)
Wald Test on Hierarchical Rank 41.93***
χ2 (df) 61.76 (6) 102.62 (10)
Log-likelihood -4,732.18 -4,672.29
N10,669 10,576
a Numbers represent odds ratios (standard errors in parentheses).
b Dummy variable coded 0 for Caucasians and 1 for other ethnicities.
† p < .10
* p < .05
** p < .01
*** p < .001.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
78
HIERARCHICAL RANK AND PRINCIPLED DISSENT
79
TABLE 3
Correlations among Variables in Study 2
Variable 1 2 3 4 5 6 7 8 9
1. High rank --
2. Low rank -.50*** --
3. Group moralitya.01 -.00 --
4. Dissent -.19** .09 -.29*** --
5. Strength of dissent -.24*** .13* -.35*** .89*** --
6. Identification .42*** -.28*** .01 -.18** -.19** --
7. Positive affect .48*** -.21*** -.09 -.09 -.12† .45*** --
8. Female gender .01 -.06 -.01 -.09 -.12* -.01 -.10 --
9. Ethnicityb.07 -.04 -.05 -.06 -.04 .15* -.02 .08 --
10. Rank manip. check .46*** -.39*** -.01 -.12* -.11† .38*** .51*** -.04 .05
a Dummy variable coded 0 for low group morality and 1 for high group morality.
b Dummy variable coded 0 for Caucasian ethnicity and 1 for other ethnicities.
† p < .10
* p < .05
** p < .01
*** p < .001
HIERARCHICAL RANK AND PRINCIPLED DISSENT
80
TABLE 4
Logistic Regression Models Predicting Dissent in Study 2a
Variable Model 1 Model 2 Model 3 Model 4 Model 5
Hierarchical Rank
High rank 0.25** (.12) 0.27* (.14) 0.25** (.12) 0.25* (.14) 0.34* (.17)
Low rank 0.97 (.36) 1.06 (.47) 0.94 (.36) 0.95 (.42) 0.89 (.34)
High group morality 0.18*** (.07) 0.21** (.12) 0.16*** (.06) 0.16** (.10) 0.15*** (.06)
Female gender 0.55† (.20) 0.55† (.20) 0.55† (.20)
Hi Rank X Group morality 0.70 (.86) 0.91 (1.15)
Lo Rank X Group morality 0.72 (.60) 0.97 (.84)
Identification 0.57* (.14)
Constant 0.64 (.18) 0.61 (.19) 0.94 (.36) 0.93 (.37) 0.84 (.33)
Wald test on rank 9.77** 7.60* 9.23** 7.45* 4.50
χ2 (df) 36.16 (3) 36.35 (5) 38.82 (4) 38.83 (6) 44.53 (5)
Log-likelihood -111.92 -111.83 -107.25 -107.25 -104.40
N266 266 264 264 260
a Numbers represent odds ratios (standard errors in parentheses).
b Dummy variable coded 0 for Caucasian ethnicity and 1 for other ethnicities.
† p < .10. * p < .05. ** p < .01. *** p < .001.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
81
TABLE 5
Means and Standard Deviations of Strength of Dissent Measure in Study 2
Condition M SD
Low Group Morality
Low Rank 3.23 1.54
Control 3.14 1.40
High Rank 2.35 1.05
High Group Morality
Low Rank 2.19 1.14
Control 2.06 1.03
High Rank 1.72 0.74
HIERARCHICAL RANK AND PRINCIPLED DISSENT
82
TABLE 6
Descriptive Statistics and Correlations among Variables in Study 3
Variable M SD 1 2 3 4 5 6 7 8
1. High rank 0.49 0.50 --
2. Identificationa0.50 0.50 -.05 --
3. Rank manipulation check 3.49 1.35 .51*** -.04 --
4. Principled dissent 0.32 0.47 -.10 .03 -.43*** --
5. Strength of dissent 2.79 1.82 -.08 .07 -.42*** .93*** --
6. Ethical views 3.58 1.42 .15†-.17*.50*** -.63*** -.64*** --
7. Female gender 0.48 0.50 -.04 -.09 -.22** .14†.14†-.16†--
8. Ethnicityb0.30 0.46 -.03 -.01 -.13 .06 .08 -.11 .01 --
9. Age 19.78 1.09 -.15†-.04 -.18*.04 .03 -.19*-.07 -.01
† p < .10. * p < .05. ** p < .01. *** p < .05
a Dummy variable coded 0 when identification with the group was impeded and 1 when it was enabled.
b Dummy variable coded 0 for Caucasian ethnicity and 1 for other ethnicities.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
83
Figure 1. Frequency of principled dissent in Study 1.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
84
Figure 2. Proportion of dissenting participants in Study 2.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
85
27%
44%
33%
20%
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
Identification Impeded Identification Enabled
Low Rank High Rank
Figure 3. Frequency of principled dissent in Study 3.
HIERARCHICAL RANK AND PRINCIPLED DISSENT
86
Appendix A
Measure of Reporting Unethical Behavior in Study 1
Did you report this activity to any of the following?
(Please mark ALL that apply.)
oI did not report the activity.
oFamily member or friend
oCoworker
oImmediate supervisor
oHigher level supervisor
oHigher level agency official
oAgency Inspector General
oOffice of Special Counsel
oLaw enforcement official
oGeneral Accounting Office
oUnion representative
oNews media
oCongressional staff member or member of Congress
oAdvocacy group outside the Government
oOther
HIERARCHICAL RANK AND PRINCIPLED DISSENT
87
Appendix B
Pre-Scripted Messages from the Group in Study 2
High Group Morality Condition
Choosing 2 is more honest. Let’s do that.
We should do what is right – msg 2.
i say we send 2. they trusted us obviously.
Message 2 is ok by me. They’d probably do the same.
Low Group Morality Condition
Choosing 1 is practical. Let’s do that.
We should do what is best for us – msg 1.
i say we send 1. we're here to make money obviously.
Message 1 is ok by me. They’d probably do the same.
