University of Utah Health Sciences Dissent and Learning Essay Dissent and LearningHow might dissent be a process and product of sensemaking? How might diss

University of Utah Health Sciences Dissent and Learning Essay Dissent and LearningHow might dissent be a process and product of sensemaking? How might dissent be desirable? What leadership communication strategies cultivate useful dissent? What might be the influence of these strategies on followers’ framing of leaders?Video source :https://www.ted.com/talks/margaret_heffernan_dare_to_disagree Also, i will upload an example Journal of Management Studies 40:6 September 2003
0022-2380
An Exploratory Study of Employee Silence:
Issues that Employees Don’t Communicate
Upward and Why*
Frances J. Milliken, Elizabeth W. Morrison and
Patricia E Hewlin
New York University
ABSTRACT There is evidence from a variety of sources that employees often do not
feel comfortable speaking to their bosses about organizational problems or issues that
concern them. The purpose of this study was to shed light on the types of issues that
employees are reluctant to raise, and identify why employees sometimes decide to
remain silent rather than voice their concerns. We interviewed 40 employees and
found that most had been in situations where they were concerned about an issue but
did not raise it to a supervisor. Silence spanned a range of organizational issues, with
several of our respondents indicating that they did not feel comfortable speaking to
those above them about any issues or concerns. The most frequently mentioned
reason for remaining silent was the fear of being viewed or labeled negatively, and as
a consequence, damaging valued relationships. From our data, we develop a model of
how the perceived consequences of voice contribute to silence, and a model of how
the social and relational implications of speaking up can take away employees’ ability
to have influence within an organizational setting.
INTRODUCTION
I raised a concern about some policies and I was told to shut up and that I was
becoming a troublemaker. I would have pursued [the issue] further but presently
I can’t afford to risk my job. This has made me go into a detached mode, making
me a ‘yes man’. (Male respondent, Information Systems)
Employees are often reluctant to share information that could be interpreted as
negative or threatening to those above them in an organizational hierarchy
Address for reprints: Frances J. Milliken, New York University, Stern School of Business, Department
of Management and Organizational Behavior, 44 West 4th Street, 7th Floor, New York, NY 10012,
USA (fmillike@stern;nyu.edu).
© BlackweU Publishing Ltd 2003. Published by BlackweU Publishing, 9600 Garsington Road, Oxford, OX4 2DQ,,
UK and 350 Main Street, Maiden, MA 02148, USA.
1454
F.J. MUlikenetal.
(Roberts and O’Reilly, 1974; Ryan and Oestreich, 1991). This reluctance to speak
up, and the silence or information withholding that it gives rise to, has the potential to undermine organizational decision making and error-correction and
to damage employee trust and morale (Argyris and Schon, 1986; Beer and
Eisenstat, 2000; Janis, 1982; Morrison and Milliken, 2000; Tamuz, 2001). Yet,
as the above quote illustrates, remaining silent may make sense when it is
considered from the perspective of the employee who fears that his or her manager
will react negatively to the upward communication of concerns and may react
further by taking steps that undermine the employee’s ability to work in the
organization.
A salient real-world example of these dynamics can be seen in the recent events
at Enron. News reports suggest that many Enron employees had concerns about
the firm’s activities but were afraid to speak to their bosses about these concerns.
According to the testimony of Sherron Watkins, a vice-president at the company,
there was ‘a culture of intimidation at Enron where there was widespread knowledge of the company’s shaky finances’, yet no one felt confident enough to raise
these issues (Oppel, 2002). This case and others provide vivid illustrations of
employee discomfort with speaking up about problems and concerns. They also
hint at the real damage that can occur (both organizationally and psychologically)
when employees feel unable to voice their concerns and therefore, do not speak
up.
Sometimes, however, employees do speak up. Sherron Watkins is one example
as is Cynthia Cooper of WorldCom and Coleen Rowley of the FBI, who were
named the 2002 People of the Year by Time Magazine for their courage in choosing to speak up to their bosses about problems they observed in their organizations. Thus, it may be most appropriate to think of the process of communicating
upward about problems or issues as the outcome of a choice that employees make
(Meyerson, 2001). Sometimes employees choose to speak and sometimes they
choose to be silent. The purpose of this research is to explore how and why
employees decide to be silent rather than voice their concerns and the types of
issues that employees feel most reluctant to raise.
Review of the Literature
One reason why people are sometimes silent about their concerns may be what
psychologists have termed the ‘mum effect’ (Rosen and Tesser, 1970). Research on
the mum effect shows that individuals have a general reluctance to convey negative information because of the discomfort associated with being the conveyer of
bad news (Conlee and Tesser, 1973). In organizations, there is evidence that
employees are especially uncomfortable conveying information about potential
problems or issues to those above them. For example, several studies have shown
that subordinates distort the information that they convey to their superiors, com© BlackweU Publishing Ltd 2003
Employee Silence
1455
municating upward in a way that minimizes negative information (e.g., Athanassiades, 1973; Read, 1962; Roberts and O’Reilly, 1974).
In other words, the hierarchical relationship between subordinate and supervisor appears to intensify the mum effect. More than half a century ago, Festinger
(1950) noted that structuring groups into hierarchies automatically introduces
restraints against free communication, particularly criticisms by low-status
members toward those in higher-status positions. Research by Athanassiades
(1973), among others, suggests that this is a form of instrumental, self-protective
behaviour. Although the evidence is inconsistent (Glauser, 1984), it appears that
employees are most likely to filter information that they convey upward when they
have high mobility aspirations (which they do not want to jeopardize) and when
they lack trust in their supervisor (Read, 1962; Roberts and O’Reilly, 1974).
In his work on organizational learning, Argyris (1977) noted that there are powerful norms and defensive routines within organizations that often prevent employees from saying what they know. Other scholars note that organizations are often
intolerant of criticism and dissent, and that employees may withhold information
in order to not ‘rock the boat’ or create conflict (Ewing, 1977; Redding, 1985;
Sprague and Rudd, 1988). As well, research that has built upon Hirshman’s (1970)
exit, voice and loyalty (EVL) model has highlighted that while voice is an option
for dealing with dissatisfying conditions, it is often not the option that is chosen
(Withey and Cooper, 1989).
More recendy, researchers have sought to explore factors that might make
people willing to communicate up hierarchies or to ‘sell’ issues to higher management. Saunders et al. (1992), for example, found that employees’ willingness to
voice work-related concerns and suggestions to their bosses depended on how
approachable and responsive they perceived their supervisors to be. This finding
is consistent with Glauser’s (1984) review of earlier works which suggested that
upward communication is affected not only by characteristics of the communicator, the message, and the organizational context, but also by characteristics of the
supervisor—subordinate relationship.
Research on issue selling has asserted that employees decide whether to raise
strategic issues with top management by ‘reading the context’ for clues concerning ‘context favourabUity’ (Ashford et al., 1998; Dutton et al., 1997, 2002). A
favourable context is described as one where top management is perceived to be
willing to listen, the culture is seen as generally supportive, and there is relatively
little uncertainty or fear of negative consequences (Dutton et al., 1997). Other
factors that have been found to affect perceived context favourability, and hence
willingness to engage in issue selling, are perceived organizational support, norms,
and the quality of one’s relationship with senior management (Ashford et al.,
1998). Although research on issue selling focuses on the decision to raise a particular strategic issue, it also suggests that employees may choose to remain silent
about issues if they conclude that the context is unfavourable.
© BlackweU Publishing Ltd 2003
1456
F.J. MUlikenetal.
An important idea from the issue selling literature is that an unfavourable
context is one where employees believe that raising an issue may lead to negative
consequences. For example, the literature has highlighted how raising a sensitive
issue such as gender inequity can damage a person’s public image. Research on
whistleblowing likewise highlights the risks (real and perceived) associated with
calling attention to problems. Whistleblowers are sometimes viewed as traitors and
can suffer negative career outcomes as a result of their calling attention to organizational wrongdoing. Research suggests that employees weigh these costs when
considering whether or not to speak up about issues and concerns (Dutton et al.,
1997; Near and Miceli, 1992).
Objectives of This Study
Despite the numerous research streams suggesting that employees often feel
uncomfortable raising issues, problems and concerns to their bosses, there is much
that we do not know about why people often remain silent. The purpose of this
study, therefore, was to see if we could better understand when and how employees decide to be silent about an issue of concern to them, and the factors that
employees consider when making this decision. Although past research provides
guidance on some of these factors, we wanted to gain a richer appreciation of how
employees think about this choice. Thus, we felt that it was most appropriate to
conduct an interview-based study in which participants shared with us their
thoughts and feelings about situations in which they had decided to remain silent
about their concerns related to issues and problems at work.
Before deciding whether to speak up about a particular issue, employees must
develop a cognitive map of the organization’s communication norms, a map of
what one can and cannot say, and of what may happen as a result of different
forms of communication. We know relatively little, however, about what these cognitive maps look like. Our main research objective was to gain a better sense of
employees’ implicit theories about speaking up or not, and how these theories
form, as a way of better understanding the decision to remain silent. In particular, we hoped that we could learn more about the types of risks and dangers that
employees are most concerned about. The issue selling literature suggests that
employees are reluctant to raise an issue if they fear that their image could be
damaged (Ashford et al., 1998). Less clear, however, is exactly what perceived
image risks entail and why employees are so concerned about damaging their
image.
In terms of how employees develop the implicit theories that may cause them
to remain silent, we were interested in seeing whether we could find evidence of
collective and social dynamics that may shape employees’ views about speaking
up. According to Morrison and Milliken’s (2000) model of organizational silence,
employees form shared beliefs about the danger and/or futility of speaking up
© BlackweU Publishing Ltd 2003
Employee Silence
1457
through processes of information sharing, social contagion, and collective sensemaking. They suggested that employees’ cognitive maps of what they can and
cannot discuss at work are shaped through observation of and communication
with others.
We were also interested in exploring the extent to which silence spans across
different issues. In other words, do employees feel that they can’t speak up at all,
or only about specific types of issues, problems, and concerns? If the latter, what
types of issues, problems and concerns are most likely to elicit silence? Studies on
issue selling, whistle-blowing, and other forms of upward communication do not
inform us much about the breadth of problems and issues that create discomfort
and that, as a result, are not communicated upward. Morrison and Milliken (2000),
however, suggested that silence might sometimes spread across a range of issues
as employees form generalized beliefs about the dangers of speaking up. Thus, a
second objective of this study was to identify the type and range of issues and concerns most likely to elicit silence.
METHODS
Sample
We interviewed 40 full time employees working in an array of industries including consulting, financial services, new media, pharmaceuticals and advertising.
Thirty per cent worked in the financial services industry, and 52 per cent were
female. Their average tenure in their current job was 4 years (min = 6 months;
max – 16 years), and 70 per cent had held prior jobs in different organizations.
We obtained the sample by soliciting students in four part-time MBA management
classes at a large urban university. We described the study as one that is focused
on how and when people choose to speak to those above them about issues or
problems that concern them at work. As an incentive to participate, we informed
the students that their names would be entered in a raffle to win a gift certificate
to one of the top-rated restaurants in the city.
Data Collection sind Interview Protocol
Each respondent was interviewed by one of the authors. The interviews lasted
2 5 ^ 5 minutes, and were conducted at the location that was most convenient for
the participants, either at school or at their work place. The interview guide
appears in Appendix A. All but two of the 40 interviews were tape-recorded.^’^
We then transcribed the tapes, focusing particular attention on the subjects’ direct
answers to our questions. We listened to each tape twice to ensure that we
captured all relevant data (i.e., all data relevant to the questions that we were
asking).
© BlackweU Publishing Ltd 2003
1458
F.J. MUliken etal.
We began each interview by providing a brief description of the study and assuring respondents that all information would be kept confidential (see Appendix A).
We then asked respondents whether, in general, they felt comfortable speaking to
their boss or to others above them about problems or issues that concern them in
their current organization. Following this general question, we asked respondents
(a) whether they could think of specific instances in their current organization
where they had felt unable to speak openly or honestly to a superior about an issue
that concerned them, and (b) whether they felt there were general classes or types
of issues that they could not raise with those above them.
Similar to other qualitative studies that have explored social dynamics at work,
we wanted to capture the richness found in personal stories (e.g., Gersick et al.,
2000). Hence, we asked respondents to tell us, in detail, about one or more situations where they felt they could not speak up about an issue of concern. We were
interested in the nature of the issue and the events surrounding the decision not
to raise it. We probed for some of the reasons why respondents chose to be silent,
either in general or about particular issues, and encouraged respondents to elaborate. In addition, whenever respondents indicated that they felt uncomfortable
raising particular issues or concerns, we asked whether they thought that others
within the organization were also uncomfortable about raising those issues. If they
said yes, we asked them to explain why they thought this. Our goal was to learn
about the collective dynamics of silence.
Content Coding
The interviews asked respondents (a) whether they were generally comfortable
expressing their opinions and concerns, (b) whether they have ever felt that they
could not openly express their opinions or concerns, (c) the issues (if any) that they
feel unable to raise at work, (d) the reasons (if any) for not raising these issues, and
(e) whether they had a basis for believing that their coworkers had remained silent
in the face of issues or problems. Respondents did not always answer the ‘yes/no’
questions with a clear-cut yes or no. Hence, there were cases where we needed to
make judgments about how to appropriately code the responses. In coding the
question about whether others in the organization were similarly uncomfortable
raising certain issues, we coded a response ‘yes’ if the respondent gave some evidence for this belief (e.g. comments that co-workers had made in private). If a
respondent said that he or she was unsure, or gave an unclear answer, we coded
this as ‘no’.
For the questions that were not coded as yes or no (e.g., the types of issues that
respondents felt they could not raise and the reasons for this reluctance), we
employed standard practices for qualitative data analysis (Miles and Huberman,
1994), engaging in an inductive process of developing and refining a coding
scheme (cf Dutton et al., 1997). To obtain an initial list of categories into which
© BlackweU Publishing Ltd 2003
Employee Silence
1459
we could code the responses, the three authors reviewed the transcripts independently and generated a list of all of the issues that respondents mentioned as well
as a list of all of the reasons that they mentioned. We then discussed and refined
the two lists, combining categories that reflected the same underlying ideas and
splitting categories where we felt that more fine-grained distinctions were needed.
Next, we conducted a preliminary coding of the data to classify the issue statements and reasons into categories (we wanted to make sure that there was not
overlap between categories or ambiguity about what type of statement should be
coded into a particular category). After several iterations, we were able to come
up with a coding checklist that we felt captured all of the issues and reasons that
respondents had mentioned (Dutton et al., 1997). Using this coding checklist, we
then went back to the 40 transcripts to code the responses, checking as many or
as few response categories as were mentioned by the respondent in the interview.
We utilized Miles and Huberman’s (1994) formula for calculating inter-rater reliability to check our coding and we obtained reliabilities that ranged from 82.5 per
cent (reasons for silence) to 99.7 per cent (whether respondent indicated having
been in a situation where he/she could not speak up). Disagreements were discussed until they were resolved (e.g., Dutton et al., 1997).
RESULTS
Our interviews suggest that being silent about issues and problems at work is a
very common experience with 85 per cent of our sample (34 of 40 people) saying
that, on at least one occasion, they had felt unable to raise an issue or concern to
their bosses even though they felt the issue was important. Only half of our respondents (51 per cent) indicated that, generally speaking, they felt comfortable speaking up about issues or concerns in their current organizations. Several of the
respondents (23 per cent of the sample) gave a definitive ‘no’ when asked whether
they generally feel comfortable speaking to their boss or to others above them
about problems or issues that concern them. Others (27 per cent of the sample)
indicated that they could speak only to certain people (e.g. their immediate boss)
or only about certain issues. In most of these cases, respondents indicated that they
could speak up about work-related issues (e.g., issues related to specific tasks or
projects) but not about more general organizational issues or concerns. Only six
respondents (15 per cent of the sample) reported that they had never felt unable
to openly express their concerns to those above them. These data are summarized
in Table I.
Issues that Respond…
Purchase answer to see full
attachment