organizational silence: a barrier to … classes/fall 07/org psy/cases... · organizational...

•'• Academy ol Management Review20CM, Vol. 25. No. 4. 70B-725.

ORGANIZATIONAL SILENCE: A BARRIER TOCHANGE AND DEVELOPMENT IN A

PLURALISTIC WORLD

ELIZABETH WOLFE MORRISONFRANCES J. MILLIKENNew York University

We argue that there are powerful forces in many organizations that cause widespreadwithholding of information about potential problems or issues by employees. We referto this collective-level phenomenon as "organizational silence." In our model weidentify contextual variables that create conditions conducive to silence and exploreihe collective sensemaking dynamics that can create the shared perception thatspeaking up is unwise. We also discuss some of the negative consequences ofsystemic silence, especially for organizations' ability to change and develop in thecontext of pluralism.

Imagine an organization where the CEO hasno clothes. The CEO's lack of clothes is apparentto all who set eyes upon him or her. Yet employ-ees never mention this. Some employees evencompliment and praise the CEO's attire. TheCEO takes pride and comfort in the fact thatsubordinates recognize his or her fine taste inclothing and easily dismisses those few trouble-makers virho look at him or her strangely or whodare to suggest that the CEO's taste in clothingis anything less than impeccable. Yet, behold,these employees are not blind. Behind the safetyof closed doors and in veiled whispers, they talkof their leader's lack of clothing. They all clearlyknow^ that the CEO is naked, but only the foolishor naive dare to speak of it in public.

Is tbis a mere fairy tale? We believe that it isnot and that far too many organizations arecaught in an apparent paradox in which mostemployees know the truth about certain issuesand problems within the organization yet darenot speak that truth to their superiors. This prop-osition can be supported from sources goingback more than 20 years, although there hasbeen little systematic exploration of why thesituation occurs. Argyris (1977), for example, dis-cussed the idea that there are powerful gamesand norms that often prevent employees fromsaying what they know about technical and pol-

We thank Peter Balsam, Danie! Beunza, lane Dutton, PatHewlin, Theresa Lant, Corey Phelps, Debra Shapiro, thieeanonymous reviewers, and special issue editor Bruce Barryfor their helpful comments on earlier drafts ot this article.

icy issues. Similarly, Redding (1985) argued thatmany organizations implicitly convey to em-ployees that they should not "rock the boat" bychallenging corporate policies or managerialprerogatives. Other scholars likewise havenoted that organizations are generally intoler-ant of dissent and that employees, thus, are re-luctant to speak up about problems {Ewing, 1977;Nemeth, 1997; Scott & Hart, 1979; Sprague &Ruud, 1988).

There are also empirical data indicating thatemployees often feel compelled to remain silentin the face of concerns or problems. It is note-worthy, however, that these data appear in pop-ular rather than academic sources, suggestingthat this topic has not received the rigorous re-search attention that it deserves. For example,in a survey of 845 line managers from diverseorganizations reported in Industry Week, only 29percent of first-level supervisors thought thattheir organization encouraged employees to ex-press opinions openly (Moskal, 1991). In addi-tion, Ryan and Oestreich (1991) conducted inter-views with 260 employees from 22 organizationsthroughout the United States and found thatmore than 70 percent indicated that they feltafraid to speak up about issues or problems thatthey encountered at work. The "undiscussables"identified in that study covered a wide range ofareas, including decision-making procedures,managerial incompetence, pay inequity, organ-izational inefficiencies, and poor organizationalperformance. The two most common reasonsthat Ryan and Oestreich's (1991) respondents

706

2000 Morrison and Milliken 707

gave for not raising these issues were that theyfeared there would be negative repercussionsfor speaking up, and they did not believe thatspeaking up would make a difference.

The possibility that the dominant choicewithin many organizations is for employees fowithhold their opinions and concerns about or-ganizational problems—a collective phenome-non that we have termed organizafionai si-ience^is one that we believe deserves seriousresearch attention. Many scholars have empha-sized the importance of upward informationflow for organizational health (Argyris & Schon,1978; Deming, 1986; Dutton & Ashford, 1993; Floyd& Wooldridge, 1994; Glauser, 1984; Saunders,Sheppard, Knight, & Roth, 1992). Scholars alsohave highlighted the importance of multipleand divergent points of view for effective organ-izational decision making {e.g., Nemeth, 1997).Thus, it seems paradoxical that so many em-ployees report feeling that they cannot commu-nicate upward about issues and problems. If isalso paradoxical that this seems to be occurringat a time when management rhetoric focuses onempowerment and more open lines of communi-cation {e.g.. Lawler, 1992; Pfeffer, 1994; Spreitzer,1995). Scholars have argued, however, that trueempowerment is not the norm in most or-ganizations (Foegen, 1999; Moskal, 1991; Pfeffer,1994).

As we will highlight, organizational silence isa potentially dangerous impediment to organi-zational change and development and is likelyfo pose a significant obstacle to the develop-ment of truly pluralistic organizations. A plural-istic organization can be defined as one thatvalues and reflects differences among employ-ees and thaf allows for the expression of multi-ple perspectives and opinions (Harquail & Cox,1993). Clearly, if people wifhin an organizationfeel that they cannot openly express their pointsof view on critical issues, then such pluralismcannot be attained. Employees may be diversein terms of their values, beliefs, priorities, andexperiences, but because expressed points ofview remain largely monolithic, the organiza-tion will not be able to benefit from this diver-sity. Thus, we argue that in order to understandhow change and development might unfoldwithin pluralistic settings, we need to under-stand the organizafional forces thaf often sys-tematically cause employees fo feel their opin-ions are not valued and that thereby discourage

them from speaking up. If is our belief fhaf onlyby tracing and understanding fhe causes of or-ganizational silence can we begin to build anaccurate and comprehensive understanding ofthe barriers to pluralism in organizations.

Our objective in fhis article is to explain boththe process by which systemic silence developswithin many organizations and fhe forces thatboth set this process in motion and reinforce if.We regard silence as a collective-level phenom-enon; hence, our model focuses on contextualrather than individual-level variables. By "con-textual" we mean factors that lie outside of theindividual actor. Our objective is not fo explainwhy a particular employee will choose to speakup or not to speak up but, rather, to explain whythe dominant response within many organiza-tions is for employees (en masse) to remain si-lent. We also integrate works from several bod-ies of literature to highlight the negativeimplications of organizational silence, particu-larly in the context of organizational changeand the quest for greater pluralism.

Before elaborating our model, we stress that itis important to recognize several theoretical andempirical works in which scholars have identi-fied antecedents to different forms of "speakingup" (Ashford, Rothbard, Piderit, & Dutton, 1998;Miceli & Near, 1992; Withey & Cooper, 1989). Spe-cifically, in research building upon Hirschman's(1970) framework, researchers have proposedand demonstrated that employees may some-times "voice" in response fo dissatisfaction (e.g.,Rusbult, Farrell, Rogers, & Mainous, 1988; Saun-ders et a l , 1992; Withey & Cooper, 1989). Therehas also been research on individual and con-textual variables that may motivate an em-ployee to try to "sell" an issue to top manage-ment (Ashford ef al., 1998) or to "blow thewhistle" in response to a perceived wrongdoing(Miceli & Near, 1992). Research has shown thatin order for employees fo express their concerns,fhey must believe that doing so will be botheffective and not too personally costly {Ashfordet aL, 1998; Miceli & Near, 1992; Withey & Coo-per, 1989).

We believe that our focus differs from theseexisting bodies of work in three ways. First, re-search on employee voice, issue selling, andwhistle-blowing has been focused on an indi-vidual employee's decision on whether or not tospeak up, whereas our focus is on collective-level dynamics. In addition, existing works (es-

708 Academy of Management Review October

pecially on issue selling and whistle-blowing)have focused on explaining why an employeemight violate the norm of silence within an or-ganization and speak up about an issue, ratherthan why such a norm exists in the first place.Yet, we believe that the forces that motivate asole individual to speak up may be quite differ-ent from the forces that compel all of the rest toremain silent. Another difference between ourwork and research on voice is that the latter hasfocused largely on individual-level antecedents(e.g., commitment, satisfaction, risk taking),whereas our focus is on the impact of factorsoutside of the individual actor. We believe thatwhen an organization is characterized by si-lence, this is less a product of multiple, uncon-nected individual choices and more a product offorces within the organization—and forces stem-ming from management—that systematicallyreinforce silence. The value of such a perspec-tive is supported by the fact that employee-levelvariables have explained relatively little vari-ance in employee voice (Withey & Cooper, 1989).

THE DEVELOPMENT OF ORGANIZATIONALSILENCE

Figure 1 provides an overview of the organi-zational conditions under which organizationalsilence is most likely to develop. Fundamen-tally, we believe that organizational silence isan outcome that owes its origins to (1) managers'fear of negative feedback and (2) a set of implicitbeliefs often held by managers. We begin byexplaining these factors, paying particular at-tention to the role of managerial beliefs. Wethen discuss the conditions under which thesebeliefs are most likely to be held by senior man-agement. Next, we discuss how managerial be-liefs and fear of feedback tend to give rise topredictable types of organizational structuresand policies, as well as to managerial practices,that impede the upward flow of information.

As Figure 1 illustrates, these structures andpractices contribute to the development of whatwe call "a climate of silence": widely sharedperceptions among employees that speaking upabout problems or issues is futile and/or danger-ous. When such a climate exists, the dominantresponse within an organization will be silence,rather than voice. However, the likelihood ofsuch a climate emerging and the strength andpervasiveness of that climate will depend on

employees' collective sensemaking activities.Hence, in the final part of this section we focuson this sensemaking process and factors thataid it.

Managers' Fear of Negative Feedback

One important factor that we believe facili-tates the creation of a climate of silence in or-ganizations is top managers' fear of receivingnegative feedback, especially from subordi-nates. There is strong evidence that people oftenfeel threatened by negative feedback, whetherthis information is about them personally orabout a course of action with which they identify(e.g.. Carver, Antonio, & Scheier, 1985; Meyer &Starke, 1982; Sachs, 1982; Swann & Read, 1981).Therefore, people try to avoid receiving negativefeedback (Ashford & Cummings, 1983), andwhen they do receive it, they may try to ignorethe message, dismiss it as inaccurate, or attackthe credibility of the source (Ilgen, Fisher, & Tay-lor, 1979).

Argyris and Schon (1978) suggest that this fearof feedback may be especially strong amongmanagers. They argue that many managers feela strong need to avoid embarrassment, threat,and feelings of vulnerability or incompetence.Hence, they will tend to avoid any informationthat might suggest weakness or that might raisequestions about current courses of action. Thereis empirical evidence that managers will be es-pecially likely to avoid negative feedback fromsubordinates. It has been shown that when neg-ative feedback comes from below, rather thanfrom above, it is seen as less accurate and le-gitimate (Ilgen et al., 1979) and as more threat-ening to one's power and credibility {Korsgaard,Roberson, & Rymph, 1998).

Managers' Implicit Beliefs

Another important factor that we believe liesat the root of organizational silence is a set ofbeliefs that managers often implicitly holdabout employees and about the nature of man-agement. One such belief is that employees areself-interested and untrustworthy. In recentworks scholars (Ghoshal & Moran, 1996; Pfeffer,1997) have emphasized that an economic para-digm currently dominates within businessschools in the United States and, hence, in thethinking of many managers. This paradigm.


o

oaSuD)a

0)01> .0

"aBtt)DlC0

D.2

u 0)0 -S

6 1- 1-.00

l ig

C C

2 0

0 uDl 3u uO tn

1̂I ^I 6O jv

Dl .«C O

U B d a,

a

EH . « - '

.2 "̂D I• s iO g

od)c

J3 tn

iSiS a0) G

w S

G 0

0) O

•c fi

0 Q

^ ^ Y\O ,G .y•u „ Sax: co -: o

C

Od3

-MUD

XI

=i G ">-P.

U Q ^ go

^ tt)

> —

O D) ^

S e c.V 0)U GC —J O

in

"s

ial

bena

ger

DU

I

icil

are

mmtt)>.n

empl

cth

atB

ell

sted

tere

!-in

i

0

cM

"50)fitt)tJ)

man

ath

at

'o)

Bel

i

T3T iGd

T3OODl

unit

yth

at

"cj

Bel

l

T lD

-QUl

II- 0110 —

D>.S

5 Dl

Is


reminiscent of what McGregor (1960) calls "The-ory X," takes as a starting assumption that indi-viduals are self-interested and act in ways tomaximize their individual utilities (Williamson,1996). In this paradigm employees are alsoviewed as effort averse, and it is argued thatthey cannot be trusted to act in the best interestsof the organization without some form of incen-tive or sanction (Ghoshal & Moran, 1996; McGre-gor, 1960; Pfeffer, 1997).

We propose that when an organization's top-level managers believe employees are self-interested and untrustworthy, they will then actin ways that implicitly and explicitly discourageupward communication. It is important to note,however, that such beliefs are generally not con-scious. In his discussions of organizationallearning, Argyris (1977) highlights that the "the-ories-in-use" that guide managers' actions dif-fer, often in significant ways, from their "es-poused theories," or what they think drives theirbehavior. Redding similarly argues that the be-liefs underlying managers' actions "usually re-main below the surface of consciousness, typi-cally unstated, only rarely being dredged upinto the light of day" (1985: 250).

A second unstated belief that we would argueis likely to create conditions conducive to organ-izational silence is the belief that managementknows best about most issues of organizationalimportance. Tbis idea bas been noted in severalsources. Ewing (1977) highlights it as the over-arching philosophy within most modern organi-zations. Glauser (1984) also notes the pervasive-ness of the belief that managers must direct andcontrol while subordinates must assume therole of unquestioning followers. Argyris (1977,1991) similarly notes that most managers be-lieve they must appear to be in unilateral con-trol, and Redding points to the implicit beliefamong managers that "hired hands should putup or shut up" (1985; 250). This belief is related tothe economic view of employees. If employeesare self-interested and effort averse, then theyare unlikely to know (or care) about what is bestfor the organization.

A third unstated belief that we regard as afactor in creating a climate conducive to wide-spread silence is the belief that unity, agree-ment, and consensus are signs of organizationalhealth, whereas disagreement and dissentshould be avoided. Burrell and Morgan (1979)describe this belief as part of the "unitary view"

of organizations, which stands in stark contrastto a "pluralistic view," in which dissent is re-garded as normal and conflict as potentiallyhealthy. The prevalence of the unitary view oforganizations also has been noted in discus-sions of group decision making, strategy formu-lation, and innovation (e.g.. Cosier & Schwenk,1990; Janis, 1982; Nemeth, 1997). Enz and Schwenk(1991) have argued that top managers hold firm tothe belief that agreement is good, despite re-search showing that the expression of multiplepoints of view relates positively to decision qual-ity and to subsequent organizational perfor-mance.

Conditions Fostering These Beliefs

A question that obviously arises concerns howpervasive the above beliefs are and the condi-tions under which they are most prevalent. Al-though we do not argue that top-level managersof all organizations hold these beliefs, the worksof several scholars indicate that they exist tosome extent in almost every organization andthat they characterize many organizations quitewell (Argyris & Schon, 1978; Enz & Schwenk,1991; Ewing, 1977; Pfeffer, 1997; Redding, 1985).For example, the assumption about employeeself-interest is likely to be common because itappears to be rooted in the system of manage-ment education currently entrenched in theUnited States (Pfeffer, 1997). Also, there is evi-dence that as people progress upward within anorganization, they are less likely to identify withsubordinates and they come to hold differentbeliefs and values from those below them(Lieberman, 1956). This decreased ability toidentify with subordinates as one climbs theorganizational hierarchy might make it easier tomake generalizations about employees (e.g.,employees are self-interested). Further, the"management knows best" belief may be solid-ified by the need to justify one's progression. Itmay be natural to assume that one must knowbest given one's superior position in the organiza-tional hierarchy.

Despite their widespread nature, however, webelieve that the managerial beliefs and as-sumptions we have described will be held morestrongly in some organizations than in others.We expect, for example, that these beliefs willbe more likely when the top management teamis dominated by individuals with economic or


financial backgrounds than when the group ismore functionally diverse or composed of indi-viduals with backgrounds in general manage-ment. Because the beliefs about employees be-ing self-interested and untrustworthy are rootedin economic models of human behavior, they aremore likely to be held by those whose trainingand job experience have been oriented towardeconomics or finance {Pfeffer, 1997), and thesebeliefs are likely to be solidified when one'speers hold them as well.

Further, the more homogeneous the top man-agement team is with respect to functionaltraining and experience, the more cohesive theyare likely to be and the more threatened theymight be by the idea of dissent (Janis, 1982; Co-sier & Schwenk, 1990). We also predict that theimpact of functional background will be inten-sified when the composition of the top manage-ment team is stable over time. The longer topmanagers have been together, the more en-trenched their shared assumptions will tend tobe and the less likely the members will be toquestion those assumptions (Wagner, Pfeffer, &O'Reilly. 1984).

Proposifion I: The managerial beliefscontributing to organizational silencewill be more common in organizationsin which the top management team isdominated by individuals with eco-nomic OI financial backgrounds.

Proposition 2: The longer the averagetenure of top management team mem-bers, the stronger will be the relation-ship between top management teamhomogeneity with respect to func-tional background and the set of be-liefs contributing to silence.

The cultural background of the top manage-ment team also may affect the beliefs that itsmembers hold about employees. For example, ifthe fop management team is composed of indi-viduals who all come from high power distancecultures, these managers especially will belikely to believe that they know best. In highpower distance cultures, individuals are muchmore likely to believe that fhe boss is rightmerely because he or she is the boss (Hofsfede,1980). Further, unity is apt to be even morestrongly valued when members of the top man-agement team are from collectivistic cultures.

since such cultures place a premium on har-mony (Hofstede, 1980). Hence, the combination ofhigh power distance and collectivism amongmembers of the top management team is partic-ularly likely to associate with silence-fosteringbeliefs.

Proposition 3: The managerial beliefscontributing to organizational silencewill be more common in organizationsin which the top management team isdominated by individuals from highpower distance and collectivistic cul-tures.

We also expect that the similarity or dissimi-larity of the demographic profile (e.g., gender,race, ethnicity, age) of fhe top managementteam in comparison to that of lower-level em-ployees might influence the prevalence of si-lence-creating beliefs. Research on diversityhas shown that salient differences often createdistrust and fear of the unknown (Cox, 1993).Hence, the greater the demographic dissimilar-ity between top management and others withinfhe organization, the more likely it is that topmanagement will view employee input withsuspicion.

Proposition 4: The managerial beliefscontributing to organizational silencewill be more common when there is ahigh level of dissimilarity (e.g., gender,race, ethnicity, age) between top man-agement and lower-level employees.

Organization-level and environmental vari-ables are also likely to affect fhe likelihood thatsilence-fostering beliefs will be held by topmanagement. For example, we predict that fhebeliefs discussed above will be more common—and, hence, silence more likely—within organi-zations dependent on predictability, control, andefficiency. When there is heavy strategic em-phasis on control, managers may view negativefeedback as more threatening and dissent asmore destructive (Bourgeois, 1985; Miceli & Near,1992). This logic suggests that a context condu-cive to silence is more likely to emerge not onlyin cost-focused organizations but also withinhighly competitive environments characterizedby a diminishing resource base. Research indi-cates that within such low-munificence environ-ments, there are greater demands for control,less openness to dissenting opinions, and often


a belief at the top that unity of views will im-prove performance (Bourgeois, 1985; Khand-walla, 1972; Pfeffer & Leblebici, 1973).

Pioposition S: The managerial beliefscontributing to organizational silencewill be more common in organizationswith a strategic focus on cost controland in organizations operating in low-munificence environments.

We also predict that the belief structure con-tributing to organizational silence will be morelikely to dominate management thought in moremature and stable industries than in newerand/or volatile industries (e.g., internet, new me-dia). In order to survive, organizations in high-velocity environments need to become adept atconsidering alternatives and responding tochanges in their environments (Eisenhardt, 1989;Lant, Milliken, & Batra, 1992). Thus, in volatileenvironmental contexts, organizations may bemore inclined to value employee ideas, sincethese ideas may be seen as useful in the searchfor new strategies (Sprague & Ruud, 1988).

Proposition 6: The managerial beliefscontributing to organizational silencewill be more common in organizationsoperating in mature and stable indus-tries.

High vertical differentiation is also likely toreinforce silence-creating beliefs. We expectthat within tall organizational structures, topmanagers will be less likely to interact with,relate to, and, hence, trust lower-level employ-ees {Blau & Scott, 1962; Glauser, 1984). In addi-tion, firms that bring in top managers from theoutside, instead of promoting from within, maybe more likely to create a gap between top man-agement and the rest of the organization. Thisgap will increase the likelihood that top manag-ers will view employees as untrustworthy or un-predictable and themselves as best equipped tounderstand and address organizational prob-lems.

Proposition 7: The managerial beliefscontributing to organizational silencewill be more common in organizationswith many hierarchical levels and inorganizations that hire senior-levelmanagers from the outside, ratherthan promote from within.

Finally, heavy reliance on contingent workersis likely to foster the managerial belief structurethat we argue contributes to organizational si-lence. Contingent employees may be viewed bymanagers as especially self-interested, ill in-formed, and, hence, untrustworthy because oftheir temporary status within the organizationand because their psychological contract withthe organization is more transactional than re-lational (Rousseau & Parks, 1993). Sennett (1998),in fact, argues that one of the consequences ofthe increasingly temporary nature of employ-ment contracts is that organizations have be-come more tightly controlled, implying a lack oftrust in employees.

Proposition 8: The managerial beliefscontributing to organizational silencewill be more common in organizationsthat rely heavily on contingent workers.

Organizational Structures, Policies, andPractices

In the prior section we discussed some im-plicit beliefs that we believe set into motion theforces that create organizational silence. In thissection we discuss how these beliefs (togetherwith managers' fear of feedback) tend to giverise to structures, policies, and managerial be-haviors that, in turn, create an environment inwhich employees feel uncomfortable speakingup about certain issues. The idea that manag-ers' beliefs and assumptions affect their behav-ior and choices, and that these then affect em-ployees, is not a new one. Forty years agoMcGregor (1960) noted that managerial beliefscan have a powerful impact on how managerstreat employees and, consequently, on how em-ployees behave (see also Argyris, 1977; Creed &Miles, 1996; Ghoshal & Moran, 1996; Pfeffer,1997). He argued, for example, that when man-agers assume that employees dislike work andcannot be trusted to do a good job, they willinstitute control mechanisms to prevent shirk-ing. Employees, realizing that managementdoes not trust them, become resentful and mayseek to find loopholes in the systems—a re-sponse that reinforces managers' initial beliefs.

In this section we propose that when the un-spoken yet dominant ideology within an organi-zation is that (1) employees are self-interested,(2) management knows best, and (3) disagree-


ment is bad, then management will erect struc-tures and policies that either do not facilitate orthat discourage upward information flow. Thistendency will be reinforced by managers' desireto avoid any threatening information or feed-back. Two common structural features of organi-zations dominated by these beliefs will be highcentralization of decision making and lack offormal upward feedback mechanisms.

If the dominant belief in an organization isthat employees are opportunistic and not knowl-edgeable about what's best for the organization,then it is reasonable for managers not to involvethem in decision-making processes (Hall, 1982).Decentralized decision making, as well as otherparticipative forms of management, will be seenby managers as not worth the time and effort theyrequire (McGregor, 1960). Excluding employeesfrom decision making is also a way to avoid dis-sent and negative feedback and, thus, will alsostem from the belief that disagreement is bad andfrom fear of feedback. Although there may be (heappearance of participative decision making (e.g.,task forces, committees), we expect true decision-making authority to reside at the top. In fact, webelieve that despite discussions of organizationsbecoming more decentralized, the implicit beliefstructure that dominates many organizationsgives rise to structures that are often quite central-ized in reality (Foegen, 1999; Sennett, 1998).

The belief structure that we believe domi-nates many organizations is also likely to beassociated with a lack of mechanisms for solic-iting employee feedback after decisions aremade. Procedures such as systematic surveyingor polling will be unlikely, because there will bea tendency to believe that little of value will belearned from them and because negative up-ward feedback will be seen as a challenge tomanagement's control. This dynamic may be aform of the threat-rigidity effect described byStaw, Sandelands, and Dutton (1981), wherebymanagement tries to protect itself from a per-ceived threat by closing itself off from input.

Proposition 9: To the extent that theimplicit belief structure of top man-agement is that employees are self-interested, management knows best,and dissent is undesirable, the organi-zation will be more likely to have cen-tralized (nonparticipatory) decision

making and will be less likely to haveformal upward feedback mechanisms.

When managers believe that employees areself-interested, opportunistic, and not well in-formed, and that agreement is preferable to dis-agreement, they also will tend to enact thesebeliefs in their day-to-day behavior toward em-ployees. For example, if employees were to ex-press concerns about a proposed organizationalchange, management would be apt to assumethat the employees were resisting the changebecause it was personally threatening to themor because they did not understand it (Kanter,1984)—not because they were truly concernedthat the change might be bad for the organiza-tion. The managers will be likely to either rejector discount opinions and feedback from employ-ees, particularly when those opinions differ fromtheir own views (Ilgen et al., 1979). Managementmay also convey, consciously or unconsciously,annoyance or even hostility toward messengersof unwanted news (Rosen & Tesser, 1970).

Finally, managers who hold these beliefs areunlikely to engage in much informal feedbackseeking from subordinates. Why seek feedbackfrom those who are self-interested and ill in-formed (Vancouver & Morrison, 1995)? And whyseek feedback from subordinates if one wishesto avoid disagreement and negative information(Ashford & Cummings, 1983)? When managersdo seek feedback, they will tend to approachthose who are likely to share their perspectiveand who are, thus, unlikely to provide negativefeedback. This will not be intentional but willsimply reflect the tendency of people to interactwith those whom they perceive to be similar tothemselves (Byrne, 1971; McPherson & Smith-Lovin, 1987) and the tendency to prefer confirma-tory feedback (Swann 8E Read, 1981).

Proposition JO; To the extent that theimplicit belief structure of top man-agement is that employees are self-interested, management knows best,and dissent is undesirable, managerswithin that organization will be morelikely to reject or react negatively toinput from subordinates and lesslikely to informally solicit feedbackfrom subordinates.

It is important to point out that the variousmanagerial practices contributing to silence

714 Academy of Managemenf flevievi' October

may operate at multiple organizational levels.Only top management has the authority to im-pose the company-wide structures and policiesthat contribute to organizational silence (e.g.,centralization). However, managers at all levelsmay exhibit the day-to-day practices fhaf im-pede upward communication (negative re-sponses to employee input, lack of feedbackseeking). This means that employees will re-ceive cues about the safety of speaking up bothfrom the fop as well as from their more immedi-afe supervisors (Ashford ef al., 1998).

We expect that in most cases, when cues fromthe top indicate that fhe organization is not openfo employee input, fhe attifudes of senior man-agement will trickle down and affect the behav-ior of middle managers (Floyd & Wooldridge,1994) and that they, foo, will send cues fo thosebelow them that voice is nof welcome. Even mid-dle managers who do not really share fhe im-plicit beliefs of fop managers may end up en-couraging silence. Recognizing that topmanagement does nof like disagreement or neg-ative feedback, middle managers may filter oufcerfain information from their subordinafes be-fore conveying fhaf informafion upward. Subor-dinafes may conclude from this behavior fhatfhe supervisor is not responsive to complaints.In other words, we expect that in most casesconsistent cues will come from middle and up-per management.

In some cases, however, employees may re-ceive cues that voice is unwelcome only fromfheir more immediate supervisor (Saunders efal., 1992). If a divisional manager discouragesand reacfs in a hosfile way fo upward commu-nication, employees within that division willfend fo coUecfively withhold their opinions, re-gardless of organization-wide policies or prac-tices. Hence, there may be widespread silenceaf fhe divisional level fhaf does not characterizefhe larger organization.

The Creation of Shared Percepfions viaCollective Sensemaking

To this point we have discussed fhe forcesfhaf lead many organizations to erect sfrucfuresand enacf pracfices fhaf impede fhe upwardflow of information—particularly informationthat indicates problems. To fully understandhow fhese structures and pracfices lead fo wide-spread withholding of informafion, or organiza-

tional silence, if is necessary fo understand fhedevelopment of what we term a climate of si-lence. If is also necessary fo undersfand fhecollecfive sensemaking process (Weick, 1995)fhrough which this climate emerges.

The term organiza(ionai climate refers foshared and enduring percepfions of psycholog-ically imporfanf aspects of a parficular workenvironmenf (Ashforth, 1985; Schneider & Reich-ers, 1983). Researchers have emphasized fhafwork settings have numerous climates, eachperfaining fo a parficular type of activity{Schneider & Reichers, 1983). For example, schol-ars have assessed organizafions' climafes forcustomer service (Schneider, 1990), innovation(Scott & Bruce, 1994), and informality (Andersson& Pearson, 1999). In addition, Miceli and Near(1992) have highlighted fhe importance of a sup-porfive climafe for whistle-blowing, althoughfhey do nof elaborafe much on fhe factors fhafmight creafe such a climafe. Researchers havealso emphasized that climate can be a morepowerful deferminanf of behavior than individ-ual needs or motivational states (Schneider &Reichers, 1983).

We define a climafe of silence as one charac-ferized by two shared beliefs: (1) speaking upabouf problems in fhe organizafion is nof worfhfhe effort, and (2) voicing one's opinions andconcerns is dangerous. This perspective mirrorswork af the individual level of analysis, whichindicates fhat employees are least likely fo en-gage in activities such as issue selling andwhistle-blowing when they judge the efficacy ofspeaking up as low and the costs of doing so ashigh (Ashford ef al., 1998 Miceli & Near, 1992;Saunders ef al., 1992; Withey & Cooper, 1989). Incontrast to work at fhe individual level, how-ever, we focus on how these judgments emergeaf a collective level and how they affecf aggre-gate-level voice behavior.

In an effort fo reconcile divergenf views ofclimafe formation, Schneider and Reichers (1983)proposed an interactionist perspective. In thisperspective it is argued that climafe originafesfrom a process of collecfive sensemaking,whereby employees together try fo derive mean-ing about fheir workplace: its demands, con-straints, and outcome contingencies (Weick,1995). The inferactionisf view of climate forma-fion builds upon a school of thought known as"symbolic interacfionism" (Blumer, 1969). Ac-cording to this school of fhought, meaning is not

2000 Moiiison and Milliken 715

a "given" but arises from interactions amongpeople (Ashforth, 1985). It is through social inter-actions that employees share their perceptionsand experiences and derive a common under-standing of the meaning of various aspects oftheir work context (Schneider & Reichers, 1983).Because individuals have a strong need to eval-uate their beliefs and perceptions (Festinger,1954), they compare them with those of others.Through this process people "triangulate on asingle set of perceptions and meanings" (Ash-forth, 1985: 839). This perspective is similar to thenotion that reality within organizations is so-cially constructed (Berger & Luckmann, 1967).

Building upon these related schools ofthought, we argue that a climate of silence isrooted not only in objective features of the work-place but also in social interactions that contrib-ute to a subjective process of sensemaking. Thestructural features and managerial practicesdiscussed in the prior section are critical, ofcourse, and can be viewed as providing the rawmaterial for silence to develop. When organiza-tional decision making is highly centralized andthere are few channels for upward communica-tion, a collective interpretation that is likely toemerge is that managers do not think employeeopinions are important (Ashford et al., 1998).When managers respond to employees' opin-ions with resistance or denial, employees aremost likely to converge on an interpretation thatspeaking up is risky or not worth the effort(Saunders et aL, 1992). The more widespreadthese structures and managerial behaviors, themore vivid and salient they will be and the morelikely they will be to foster shared perceptionsthat speaking up is not welcome. However, thelikelihood that these structures and actions willlead to a climate of silence will be greater whenthe context provides opportunities for employeesto form common perceptions. As shown in Figure1, we propose that a process of collective sense-making moderates the relationship between or-ganizational structures and managerial practiceson the one hand and climate on the other. Hence,to understand how a climate of silence developswithin an organization, it is necessary to under-stand this sensemaking process and the factorsthat make it more likely to occur.

The literature on symbolic interactionism, aswell as the social network literature, indicatesthat common perceptions and attitudes are mostlikely to develop and become reinforced to the

extent that members of a social unit have oppor-tunities to interact and communicate with oneanother (Erickson, 1988; Ibarra & Andrews, 1993;Meyer, 1994). The more frequent and intense thecontact and communication, the more likely it isthat "social contagion"—a spreading of at-titudes and perceptions from one person to an-other—will occur (Erickson, 1988; Rice & Aydin,1991). Social network research has indicatedthat the following variables facilitate contactand communication and, hence, the develop-ment of common perceptions: similarity, stabil-ity, workflow interdependence, and informal so-cial networks that are dense and composed ofstrong ties. Thus, assuming there are structuraland managerial factors conducive to organiza-tional silence in place (e.g., centralized decisionmaking, lack of upward feedback mechanisms,defensive reactions to employee input), we ex-pect that similarity, stability, workflow inter-dependence, and the strength and density ofsocial network ties among mid- to lower-levelemployees will increase the likelihood of astrong climate of silence developing.

It has been found that individuals are mostcomfortable and prefer to interact with thosewhom they perceive to be similar to themselves(Byrne, 1971; Ibarra, 1992). Individuals also pre-fer similar others as referents for validatingtheir beliefs and perceptions (Festinger, 1954).Hence, given an organizational context with fewmechanisms for employee voice, there will be agreater likelihood of a climate of silence emerg-ing when there is a high degree of similaritybetween employees. Similarity between directcoworkers will facilitate social contagion at theworkgroup or departmental level, while bases ofsimilarity between individuals working in dif-ferent areas will foster information sharing andsocial contagion from one area to another.Shared perceptions are also more likely to de-velop within organizations with relatively stablemembership, since stability implies more oppor-tunities for employees to interact and share theirperceptions and a greater likelihood of sharedperceptions enduring over time.

Workflow interdependence is another vari-able that will contribute to the collective sense-making process, which, in the presence of cer-tain structural features, may give rise to aclimate of silence. When employees within aworkgroup must coordinate their activities,there is greater necessity for communication

716 Academy oi Management Review October

and, thus, a greater opporfunify for fhem foshare their percepfions and experiences (Meyer,1994). Further, when employees musf coordinatefheir acfivifies across workgroups or areas (e.g.,cross-functional teams), fhere is greafer oppor-tunity for informafion to be exchanged and forpercepfions to be shared between workgroups.Hence, as employees attempt fo make sense ofsuch sfructural features as fop-down decisionmaking and closed feedback channels, they willbe more likely to converge on and reinforce fheinterpretation fhaf fhe organization is hosfile totheir inpuf when fheir work necessitates regularcommunicafion, coordinafion, and teamwork.

In addifion fo formal sfructural arrangemenfs,employees are also linked fo one another viainformal social networks. Ibarra and Andrews(1993) highlight fhaf informal nefwork links arecritical channels through which organizafion-ally relevant beliefs and aftitudes are commu-nicated. A social network is composed of strongties when fhe linkages wifhin fhaf nefwork rep-resenf frequenf and intense contact (Granovet-ter, 1973). Because strong ties imply more inter-acfion and communicafion, fhey imply moreopporfunity for collective sensemaking and agreafer likelihood that employees will developshared percepfions abouf the efficacy and riskof speaking up. Network density is also likely fofacilitate collective sensemaking. A dense nef-work is one with few missing linkages, implyingfhaf a given individual is likely to receive infor-mation from many differenf sources yet largelyfile same information from these differentsources (Burf, 1992). The intense and consistentcommunication that occurs wifhin a dense so-cial network suggests that more collecfive sense-making will take place.

Proposition 11: Centralized decisionmaking, lack of upward feedbackmechanisms, managerial resistanceto employee input, and lack of down-ward feedback seeking will he morelikely to lead to a climate of silence tothe extent that mid- to lower-level em-ployees directly interact and commu-nicate among themselves. •

Proposition 12: The amount of interac-tion and communication among mid-to lower-level employees will be pos-itively related to social similarity,workforce stability, workflow inter-

dependence, and the strength anddensity of the informal social net-works connecting those employees.

Distortions in the Sensemaking Process

In the above discussion we elaborated on fhecollecfive sensemaking process fhaf can impacffhe emergence of a climafe of silence. An addi-tional point worth highlighf ing is that this sense-making process has a strong tendency fo giverise to biased and often inaccurate percepfionsas employees share and coUecfively inferpreffheir observations and experiences (Ashforth,1985). Employees make sense of managerial ac-fions based on limited and offen distorted infor-mation. Much of this information is secondhand,because people prefer fo learn vicariously abouffhe types of behaviors fhaf are dangerous ratherthan risk finding ouf firsthand. If is far safer, forexample, to believe a colleague who says fhafspeaking up will earn one fhe repufation of be-ing a "troublemaker" than if is to test fhe sysfemand risk losing one's accumulated credibility.

Indeed, research indicates fhat individuals of-fen give more credence fo what others seem fobelieve fhan what fhey themselves perceive fobe true (Nemeth, 1997). So even if one's personalexperience suggests that voice is welcomed, ifothers seem to regard fhe organizafion as unre-ceptive fo voice, fhen fhe individual will fend foadopt this view as well. Furfher, ofhers wifhinthe organizafion choosing fo remain silent aboutissues and problems will reinforce fhe percep-tion thaf speaking up is taboo. This is illusfrafedby fhe feedback loop in Figure I.

The process of collective sensemaking is alsolikely to give rise to exaggerated percepfions ofthe riskiness and futility of speaking up. Salienfevenfs offen appear fo have a larger causal im-pact than fhey actually do (Kiesler & SprouU,1982). Thus, if a member of the organizationvoices dissenf and soon thereafter fails to re-ceive an expected promotion, some employeesmay reach fhe conclusion that fhe promofionwas lost because fhis person expressed an un-popular opinion. As fhis information passesfhrough the grapevine, there may soon arise thewidespread perception that fhose who expressnegative views are punished. Similarly, if a fewemployees offer inpuf on a parficular policychange and that inpuf is ignored, fhey may con-clude fhat all input is ignored, even if fhis is nof


the case. As fhese employees convey fheir inter-pretation fo coworkers, fhe laffer, too, may cometo believe fhaf voice is nof worfh fhe effort, eventhough they lack any firsthand evidence to sup-porf this belief.

Shared beliefs about speaking up being de-valued or dangerous are likely to become notonly exaggerated buf also overgeneralized. Thatis, silence may af firsf center on certain issues,buf eventually it may spread far beyond thoseissues, and employees may sfop sending anyinformation up fhe hierarchy. Greenberger andSfrasser's (1991) work supports this assertion.They suggesf fhaf when people perceive a lossof control in one situation (e.g., a suggestion thatfhey make is ignored), this may carry over andaffecf fheir behavior in ofher sifuations. Ash-forfh similarly notes fhis tendency fo overgener-alize, remarking how "a workgroup may come toagree fhaf a seemingly positive overture fromfheir supervisor has an ulterior mofive andeventually develop the belief fhaf any overturefrom any member of managemenf is suspect"(1985: 840). Such overgeneralized beliefs willalso be resisfanf to change, because individualsdevelop a common stake in fheir perpetuation(Woodman & King, 1978).

A naive manager, were he or she aware ofemployees' exaggerafed and overgeneralizedperceptions abouf speaking up, mighf dismissthem as inaccurate. It is well known, however,fhaf perceptions, regardless of whether they areaccurate, have a strong impact on subsequentattitudes and behavior, and the impact of per-ceptions and beliefs is especially strong whenthey are widely shared and supported by socialinformation from peers (Salancik & Pfeffer, 1978).The outspoken individual in the example citedabove mighf have losf fhe promofion for reasonsunrelated fo the prior expression of dissent. Yet,as long as employees perceive the fwo events asrelated, fheir shared perception that speakingup is dangerous will be reinforced, and they willbe more likely fo remain silenf fhemselves.

Direcf Effecfs of Demographic Dissimilarity

Demographic dissimilarity between top man-agers and lower-level employees was a factorthat we hypothesized would increase the likeli-hood of management holding beliefs that con-tribute to silence. This variable is also likelyto contribute more directly to a climate of si-

lence by affecting fhe perceptions and beliefs oflower-level employees. Research has shownthat the common experience of being differentfrom those in positions of power leads fo somepredicfable reacfions on fhe parf of those atlower levels in the hierarchy (Ely, 1994). When alarge number of employees see that people likethemselves are underrepresented at the top,they may be more likely to conclude that fheorganization does not value the input of peoplelike themselves. This conclusion, in turn, mayfoster the belief that it is even more risky forthem to honestly voice their opinions than it isfor employees who are more similar to those atthe top.

Kanter speculates that people who perceivethemselves as having little or no opportunity foradvancement will be "less likely to profesf di-rectly or seek change" and more likely to "chan-nel grievances into griping and output resfric-fions rather than direct acfion" (1977: 247). Ourargument is also consisfenf with fhe social nef-work literafure, which has shown that individu-als are more likely to have similar perceptionsand atfifudes when they hold comparable posi-tions within the organizafion or view fhe or-ganizafion from a similar vantage point (Meyer,1994; Rice & Aydin, 1991). The "comparable posi-tion" in this case is one of exclusion.

Proposition 13: A climate of silencewill be more likely when there is ahigh degree of demographic dissimi-larity between top managers and em-ployees below the level of top man-agement.

IMPLICATIONS OF SILENCE FORORGANIZATIONS

Thus far, we have presented a model of howsilence unfolds within organizations and haveidentified factors that increase the likelihood itwill develop. In this section we briefly turn ourattention to the organizational implications ofsilence. Our objective is not to provide an ex-haustive discussion but, rather, to integrate ex-isting research in a way that highlights some ofthe ways in which silence mighf affecf criticalorganizational processes and outcomes. Wealso note how variables such as diversity anddemographic dissimilarity might make thesenegative effects more severe. Figure 2 providesan overview of these ideas.

718 Academy oi Management Review October

>0)

*-"0)cncn0)

DCo0N

'S0

o

u)d)OJUlQ)U01M

a03O)

cax lu

cn

CM

U

01—i

0

"5N'S0aO0•it

[ect

f

S3

• ^

u01

01inin

'^

co

u0

T )

c0dN

'Sdd)0

0]G• J

dS

rror

0)

0o

Cd

ion

tect

CXTJ

tion

u01

o

ud'o

II*H d 'M

ud)_o

6 M

y

a0

o

s01

o

> •

dILU

IS

cn

• - •

o'5"Ql

J 3


Effects on Organizational Decision Making andChange Processes

One significant effect of organizational si-lence relates to organizational decision makingand change, as highlighted in the top portion ofFigure 2. Extensive research on group decisionmaking has shown that decision quality is en-hanced when multiple perspectives and alter-natives are considered (see Shaw, 1981). Simi-larly, research on strategy formulation hasshown that the existence of multiple and con-flicting viewpoints within top managementteams has a positive effect on both the quality oforganizational decision making and companyperformance {Bourgeois, 1985; Enz & Schwenk,1991). Further, it has been argued that innova-tion requires a context in which employees feelfree to deviate—to offer totally novel perspec-tives or ideas or to question current beliefs andpractices (Nemeth, 1997). Together, these re-search streams suggest that organizational si-lence will compromise the effectiveness of or-ganizational decision making and changeprocesses by restricting the variance in informa-tional input available to decision makers. Inaddition, without minority viewpoints there isless likely to be the type of critical analysisnecessary for effective decision making, whichmay also compromise organizational changeprocesses (Nemeth, 1985; Nemeth & Wachter,1983; Shaw, 1981).

Of course, we are not claiming that unlimitedemployee input is desirable. Too much inputmight overload decision-making processes andimpede timely and effective decision making(Glauser, 1984). What we do claim though is thatmost organizations suffer from (oo little em-ployee voice, particularly around problems ordecisions that employees perceive to be unwise(Argyris, 1977; Ryan & Oestreich, 1991; Scott &Hart, 1979). When this is the case, the quality ofdecision making and change initiatives is likelyto be lessened.

Another way in which organizational silenceis likely to compromise effective organizationalchange and development is by blocking nega-tive feedback and, hence, an organization's abil-ity to detect and correct errors {Miller, 1972).Without negative feedback, errors tend to per-sist and may even intensify, because correctiveactions are not taken when needed (Agyris &

Schon, 1978; Deming, 1986; March, 1991). Silenceimplies that an organization will lack the capac-ity for what Argyris (1977) calls "double-loopedlearning," which entails a questioning and mod-ification of underlying policies and goals.

To make matters worse, top managers maynot recognize that they are lacking importantinformation and may interpret silence as signal-ing consensus and success. Even if managersdirectly ask employees for feedback, employeesmay be careful to filter out negative information.As a result, the internal feedback that manage-ment receives may reflect what employees thinkmanagement wants to hear, rather than accu-rate or reliable reactions (Zand, 1972).

In a recent study Bies and Tripp (1999) foundevidence that fear of repercussions does indeedcause employees to distort the information theyprovide to their bosses; these authors make thefollowing observation: "Bosses, based on theirobservations of the employees' 'loyal' and obe-dient behavior, made inferences that led to afalse consciousness and a false consensus as tothe level of affection or disaffection with theirleadership" {1999: 17). If a manager uses suchdistorted feedback as a basis for decidingwhether to take further action, the organizationruns the risk of drifting farther and farther offtarget {Argyris, 1977; Miller, 1972).

These dynamics are apt to be particularlyproblematic in organizations in which the es-poused goal is to become more pluralistic. Plu-ralism implies tolerance of multiple and oftenconflicting viewpoints. Yet, when the system re-inforces silence, the diverse viewpoints, opin-ions, preferences, and goals that may existwithin the organization are not likely to be givenvoice and, hence, will not enter into the pro-cesses by which the organization establishesobjectives, decides on appropriate courses ofaction, and attempts to learn from experience. Aproposition that we would offer, in fact, is thatthe negative effects of silence on organizationaldecision making and change processes will beintensified as the level of diversity within theorganization increases. We would also proposethat organizational silence will be most detri-mental within rapidly changing environments.Within such environments it is virtually impos-sible for those at the top to have all of the infor-mation they need {Duncan & Weiss, 1979).

720 Academy o/ Management Review October

Effects on Employee Cognitions, Attitudes, andBehavior

If you experience fear every day. it drags youdown and you become cowardly.

After my suggestions were ignored, the quality ofmy work was still there, but I wasn't.

(Ryan & Oestreich, 1991: 7)

The tendency of organizations to discourageemployee opinions and feedback is not onlylikely to compromise organizational decisionmaking and change but is also likely to elicitundesirable reactions from employees. Drawingfrom a variety of literature, we propose thatthree destructive outcomes of organizational si-lence will be employees who feel they are nofvalued, employees who perceive they lack con-trol, and employees who experience cognitivedissonance. As shown in Figure 2, these threeoutcomes are likely to lead to a variety of sec-ond-order outcomes that, in the aggregate, mayundermine organizational effectiveness.

Employees' feelings of not being valued. Re-search on procedural justice has consistentlyshown that employees evaluate decision proce-dures more favorably when those proceduresallow for employee input, even when this inputdoes not have much impact on decision out-comes (Bies & Shapiro, 1988; Lind, Lissak, & Con-Ion, 1983; Tyler, Rasinski, & Spodick, 1985). Ac-cording to Lind and Tyler's (1988) group valuemodel, procedures that allow for employee voiceare viewed positively, at least in part, becausethey signal that employees are valued membersof the organization. The model also suggeststhat employees feel unvalued when they per-ceive that they and others cannot openly ex-press their viewpoints. Eisenberger, Fasolo, andDavis-LaMastro's (1990) work indicates thatthese feelings will affect organizational commit-ment and trust. If employees feel their organiza-tion does not value them, they will be less likelyto value, identify with, or trust the organization.Outcomes that may follow from diminishedcommitment and trust are lower internal moti-vation and satisfaction, psychological with-drawal, and even turnover {Mathieu & Zajac,1990).

Employees' perceived lack of control. There isevidence from several sources that individualshave a strong need for control over their imme-diate environment and over decisions that affectthem (see Greenberger & Strasser, 1986; Parker,

1993; Wortman & Brehm, 1975). Research on pro-cedural justice indicates that an important wayin which employees gain a sense of control overtheir environment is by expressing their opin-ions and preferences (Lind & Tyler, 1988). Thiswork further indicates that employees' need forcontrol is unmet when they are denied the op-portunity to voice {Lind & Tyler, 1988). Hence, wecan expect that when employees are surroundedby social cues that discourage speaking up, theywill feel they lack sufficient control over theirwork environment.

Evidence exists that a felt lack of control hasseveral detrimental effects, including reducedmotivation, dissatisfaction, stress-related ail-ments, physical and psychological withdrawal,and even sabotage or other forms of deviance(see Greenberger & Strasser, 1986; Parker, 1993).Outcomes such as sabotage may reflect "reac-tance" or an attempt to regain control (Brehm,1966). If employees feel that they cannot exertcontrol through voice or other constructivemeans, they might try to demonstrate control inways that are more destructive for the organiza-tion (Ashforth & Lee, 1990; Folkman & Lazarus,1980; Greenberger & Strasser, 1986). Conversely,outcomes such as stress and withdrawal mightreflect a learned helplessness response (Selig-man, 1975). It is worth noting that these variousresponses may serve to validate managers' ini-tial assumptions that created a climate of si-lence in the first place. Employees are apt toappear as either hostile opponents who cannotbe trusted or as apathetic observers who areunwilling to contribute much beyond what theyare required to contribute. In essence, manag-ers' beliefs may become self-fulfilling {McGre-gor, 1960; Pfeffer, 1997).

Employees' cognitive dissonance. Organiza-tional silence is also likely to give rise to cogni-tive dissonance: an aversive state that ariseswhen there is a discrepancy between one'sbeliefs and one's behavior {Festinger, 1957). In-dividuals experiencing dissonance are typicallymotivated to try to restore consistency by chang-ing either their beliefs or their behavior. Yet, inthe context of organizational silence, it may bevery difficult for employees to reduce dissonance.

Consider a salesperson who is confronteddaily with evidence that customers are not sat-isfied with a particular product. This salesper-son may have a very difficult time believingmanagement's pronouncements that the product

2000 Morrison and Miliiken 721

is great. But the other option for reducing disso-nance^speaking up about the product's flaws—may be highly risky.

In cases such as this, when neither beliefs norbehaviors can be easily changed, the individualmay exist in a state of prolonged dissonance,which is likely to give rise to anxiety and stress{Parker, 1993; Pennebaker, 1985). The ability tovent to trusted colleagues might reduce but willnot eliminate the dissonance, since there will bea stark contrast between what one expressesbehind closed doors and what one expresses {orfails to express) in public {Bies & Tripp, 1999;Morris & Feldman, 1996).

The above dynamics are particularly trou-bling, because they may disproportionately af-fect those who differ from the majority and whotherefore are most likely to have different per-spectives. Not only will such employees feelgreater pressure to remain silent {as discussed),but they also may be more likely to experiencethe negative effects of silence. In other words,we predict that the relationship between silenceand the three psychological outcomes will bemoderated by demographic dissimilarity {seeFigure 2). If this prediction is true, it may be animportant factor contributing to systemic, yetunintentional, discrimination in organizations.

DISCUSSION AND CONCLUSION

In the call for papers for this special issue ofthe Academy of Management Review, it is ar-gued that "if the expression of opposition is notwelcomed in organizations, then new theoriesand models that allow for the existence andeven the encouragement of divergent view-points and expressions are needed." Yet thismay not be so simple. We have argued thatthere are powerful forces operating in organiza-tions that can make employees feel that speak-ing up about issues and problems is futile or,worse yet, dangerous. We see the silence thatresults as a critical barrier to organizationalchange and development, as well as a signifi-cant demoralizing force. Our focus, thus, has notbeen on how change and development occur {orshould occur) in pluralistic settings but on whyorganizational change often fails to capitalizeon or reflect pluralism. Capitalizing on plural-ism requires creating systems that enable em-ployees to voice. Effectively creating systemsthat encourage voice, however, requires an un-

derstanding of the complex dynamics withinmany organizational systems that maintain andreinforce silence instead.

Our model indicates that when top manage-ment adheres to the assumptions that foster si-lence, it may be especially difficult for the or-ganization to respond appropriately to thediversity of values, beliefs, and characteristicsthat increasingly characterize the workforce.The more these differences "pull" the organiza-tion in divergent directions, the more the or-ganization will "push" against or resist theseforces. This tendency will not be conscious butwill be driven by the implicit set of assumptionsthat govern the organization. If employees areself-interested and dissent bad, then differentviewpoints are a threat to be suppressed.

A troubling aspect of the dynamics that createand maintain silence is that they are hiddenfrom view. It is like the tale of the emperor's newclothes. Everyone understands that it is risky tospeak the truth, but this fact itself is "undiscuss-able," and few people know what to do to bringabout change. Although the hidden dynamicsthat create silence can create many problemsfor the organization, the reasons for these out-comes are unlikely to be understood. Manage-ment may see that employees are not engagedbut may assume that it is because they are self-interested and opportunistic. In addition, withinorganizations plagued by silence, problemsmay accumulate to the point that they can nolonger be hidden from the view of importantstakeholders, such as owners or creditors. Atthis point, these constituencies may concludethat the organization suffers from "poor man-agement," and top managers may lose theirjobs.

Another troubling issue is the possibility ofemployees implicitly being blamed for their ten-dency to remain silent in the face of organiza-tional problems. If the powerful systemic forcesthat create organizational silence are not recog-nized, there is the danger of holding employeesaccountable for something that really stemsfrom the actions of managers. We as researchersmay even be guilty of this, particularly when werely on models of voice that emphasize individ-ual-level determinants. Just as Whitener, Brodt,Korsgaard, and Werner (1998) argue that it ismanagers' responsibility to create trust withintheir institutions, we argue that it is the respon-sibility of management to break down (or never


erect) the conditions that sustain silence withinmany organizations, and it is the role of re-searchers to help them understand how to dothis.

We believe, however, that the behavioral cy-cles that maintain organizational silence will beextremely hard to break. One reason is thatthese cycles are not subject to direct observationor discussion. Another is that once people startdistrusting a system, it is extremely hard to re-store their faith {Zand, 1972). Even if managerseventually realize that they need accurate inter-nal feedback and try to elicit it, employees willtend to be cynical about this change (Reichers,Wanous, & Austin, 1997). Cynicism, like distrust,may be difficult to eliminate once it takes root.

Changing an organization from one charac-terized by a climate of silence to one that en-courages voice may require revolutionarychange to the system (Tushman & Romanelli,1985). Yet, achieving such system-wide changeis unusual in the absence of a strong externalforce. Even organizations faced with informationthat their performance has declined will notnecessarily engage in strategic reorientations(Lant et al., 1992). Top management change willbe a necessary but insufficient condition for theorganization to break out of a climate of silence.Theoretically, senior managers with differentdriving assumptions ought to be able to designdifferent types of organizational systems that,over time, will be successful in eliciting honestupward communication. However, it will takeenormous time and effort to overcome employ-ees' widely shared and deeply imbedded per-ceptions that speaking up is unwise. In order tochange these perceptions, it may be necessaryfor top management turnover to be accompa-nied by substantial turnover at multiple levelsthroughout the organization.

In conclusion, it has been common for sometime now for both scholars and practitioners toespouse goals related to greater empowerment,employee involvement, and pluralism. Yet, inspite of this rhetoric, there seems to be compel-ling anecdotal evidence (Argyris, 1977; Moskal,1991; Redding, 1985; Ryan & Oestreich, 1991) thatit is common for employees to think that speak-ing up is useless and even dangerous—beliefsthat are clearly at odds with notions of empow-erment, involvement, and pluralism. In this ar-ticle we have tried to shed some light on thisparadox. We have tried to explain why, despite

"knowing" that they should encourage upwardcommunication, organizations' dominant ten-dency may be just the opposite—namely, to cre-ate a climate of silence. We have tried to explainwhy true pluralism, for most organizations, re-mains elusive. And we have tried to explain whychange and development in the context of plu-ralism require not just alternative models andtheories but also a recognition of the forces thatmay make it difficult, and even impossible insome contexts, to apply those models and theo-ries.

REFERENCES

Andersson, L. M.. & Pearson, C. M. 1999. Tit ior tat? Thespiraling effect of Incivility in the workplace. Academyof Management Review, 24: 452-471.

Argyris, C. 1977. Double loop learning in organizations. Har-vard Business Review, 55(5): 115-129.

Argyris, C. 1991. Teaching smart people how to learn. Har-vard Business Review, 69(3): 99-101.

Argyris. C. & Schon, D. 1978. Organizational learning. Read-ing, MA: Addison-Wesley.

Ashiord, S. J., & Cummings, L, L, 1983. Feedback as an indi-vidual resource: Personal strategies oi creating iniorma-tion. Organizational Behavior and Human Performance,32: 370-398.

Ashford, S. J., Rothbard, N. P., Piderit, S, K., & Dutton. J. E.1998, Out on a limb: The role oi context and impressionmanagement in selling gender-equity issues. Adminis-trative Science Quarterly, 43: 23-57.

Ashforth, B. E, 1985, Climate iormation: Issues and exten-sions. Academy ol Management Review, 10: 837-847.

Ashforth, B, E., & Lee, R. T. 1990. Defensive behaviors inorganizations: A preliminary model. Human Relations.43: 621-648,

Berger, P, L., & Luckmann, T. 1967. The social construction o/reality. Hevj York: Doubleday.

Bies, R. J., & Shapiro, D, L, 1988, Voice and justification: Theirinfluence on procedural fairness judgements. Academyoi Management Journal, 31: 676-685.

Bies, R, I,, & Tripp, T. M. 1998, Two iaces of the powerless:Coping with tyranny. In R, M, Kramer & M, A, Neale(Eds,), Power and inlluence in organizations: 203-219.Thousand Oaks, CA: Sage.

Blau, P. M., & Scott, R, 1962. Formal organizations. San Fran-cisco; Chandler Publishing,

Blumer, H. 1969. Symbolic interacfionism: Perspective andmethod. Englewood Clifis, NJ: Prentice-Hall.

Bourgeois, L. 1985, Strategic goals, perceived uncertainty,and economic performance in volatile environments.Academy ol Management Journal, 28: 548-573.

Brehm, J. W. 1966. Response to loss ot freedom: A theory olpsychological reactance. New York: Academic Press.


Burrell, G., & Morgan, G. 1979. Sociological paradigms andorganizational analysis. London: Heinemann.

Burt, R. S. 1992. Structural holes: The social structure of com-pefition. Cambridge, MA: Harvard Universiiy Press.

Byrne, D. 1971. The attraction paradigm. New York: AcademicPress.

Carver, C. S., Antonio, M., & Scheier, M. F. 1985. Seli-consciousness and self-assessment. Journal oi Person-ality and Social Psychology. 48: 117-124.

Cosier, R. A., & Schwenk. C. R. 1990. Agreement and thinkingalike: Ingredients ior poor decisions. Academy of Man-agement Executive. 4(1): 69-74.

Cox, T. 1993. CuJturai diversity in organizations: Theory,research, and practice. San Francisco: Berrett-Koehler.

Creed, W. E. D., & Miles, R. E. 1996. Trust in organizalions: Aconceptual framework linking organizational lorms,managerial philosophies, and the opportunity costs ofcontrols. In R. M. Kramer & T. R. Tyler (Eds.), Trust inorganizations: Frontiers of theory and research: 16-38.Thousand Oaks, CA: Sage.

Deming, W. E. 1986. Out of the crisis. Cambridge, MA: MITPress.

Duncan, R., & Weiss, A. 1979. Organizational learning: Im-plications for organizational design. In B. M. Staw (Ed.),flesearch in organizarionaj behavior, vol. 1: 75-123.Greenwich, CT: JAI Press.

Dutton. J. E., & Ashlord, S, ]. 1993. Selling issues to top man-agement. Academy of Management BeWew, 18: 397-428.

Eisenberger, R., Fasolo, P., & Davis-LaMastro, V. 1990. Per-ceived organizational support and employee diligence,commitment, and innovation. Journal oi Applied Psy-chology, 75: 51-59.

Eisenhardt, K. 1989. Making fast strategic decisions in high-velocity environments. Academy oi Management Jour-nal 32: 543-578.

Ely. R. J. 1994. The effects of organizational demographicsand social identity on relationships among professionalwomen. Administra(ive Science Quarterly, 39: 203-238.

Enz, C. A., & Schwenk, C. R. 1991, The performance edge:Strategic and value dissensus. Employee Responsihilityand Rights Journal. 4: 75-85.

Erickson. B. H. 1988. The relational basis oi attitudes. InB. Wellman & S. Berkowitz (Eds.), SociaJ structures: Anetwork approach: 99-121. Cambridge: Cambridge Uni-versity Press.

Ewing, D. W. 1977. Freedom inside the organization. NewYork: Dutton.

Festinger, L. 1954. A theory of social comparison processes.Human fleiations, 40: 427-448.

Festinger, L. 1957. A theory o/cognitive dissonance. Stanford,CA: Stanford University Press.

Floyd, S. W., & Wooldridge, W. 1994. Dinosaurs or dynamos?Recognizing middle management's strategic role. Acad-emy of Management Executive. 8(4): 47-57.

Foegen, J. H. 1999. Why not empowerment? Business andEconomic Review, 45(3J: 31-33.

Folkman, S., & Lazarus, R. S. 1980. An analysis of coping in amiddle-aged community sample. Journal oi Health andSocial Behavior. 21: 219-239.

Ghoshal, S., & Moran, P. 1996. Bad for practice: A critique ofthe transaction cost theory. Academy of ManagementJournal 21: 13-47.

Glauser, M. J. 1984. Upward information flow in organiza-tions: Review and conceptual analysis. Human Rela-tions, 37: 613-643.

Granovetter, M. 1973. The strength of weak ties. AmericanJournal oi Sociology. 78: 1360-1380.

Greenberger, D. B,, & Strasser. S. 1986. The development andapplication of a model oi personal control in organiza-tions. Academy of Management Review. 11: 164-177.

Greenberger. D. B., & Strasser, S. 1991. The role of situationaland dispositionai factors in the enhancement of per-sonal control in organizations. In L. L. Cummings & B. M.Staw (Eds.), Research in organizational behavior, vol. 13:111-145. Greenwich, CT: JAI Press.

Hall, R. H. 1982. Organizations; Structure and process. Engle-wood Cliffs, NJ: Prentice-Hall.

Harquail. C. V., & Cox, T. 1993. Organizational culture andacculturation. In T. Cox, Jr. (Ed.), Cultural diversity inorganizations: 161-176. San Francisco: Berrett-Koehier.

Hirschman, A, O. 1970. Exit, voice, and loyalty: Responses todecJine in firms, organizations, and states, Cambridge,MA: Harvard University Press.

Hoistede, G. 1980. Culture's consequences: International dif-ferences in work-related values. Beverly Hills, CA: Sage.

Ibarra, H. 1992. Homophily and diiferential returns: Sex dif-ferences in network structure and access in an advertis-ing firm. Administrative Science OuarterJy, 37: 422-447.

Ibarra. H., & Andrews, S. B. 1993. Power, social influence, andsense making: Effects of network centrality and proxim-ity on employee perceptions. Administrative ScienceQuarterly. 38: 277-303.

Ilgen, D. R., Fisher, C. D.. & Taylor, M. S. 1979. Consequencesoi individual feedback on behavior in organizations.Journal of Applied Psychology. 64: 349-371.

Janis, I. L. 1982. Groupthini. Boston: Houghton Mifilin.

Kanter, R. M. 1977. Men and women of the corporation. NewYork: Harper 8E ROW.

Kanter, R. M. 1984. Managing change: The human dimension.New York: Harper & Row.

Khandwalla, P. N. 1972. Environment and its impact on theorganization, /nfernationai Studies of Management andOrganization. 2: 297-313.

Kiesler, S., & SprouU. L. 1982. Managerial responses tochanging environments: Perspectives on problem sen-sing from social cognition. Administrative ScienceQuarterly, 27: 548-570.

Korsgaard, M.. Roberson. L., & Rymph, R. D. 1998. What mo-tivates fairness: The role of subordinate assertive be-havior on managers' interactional fairness. Journal ofApplied Psychology. 83: 731-744.

Lant. T.. Milliken, F.. & Batra, B. 1992. The role of managerial

724 Academy ot Management Review October

learning and interpretation in strategic persistence andreorientation: An empirical exploration. Sfrofegic Man-agement Journal, 13: 585-608,

Lawler, E, E, 1992. The ultimate advantage: Creating thehigh-involvement organization. San Francisco: Jossey-Bass.

Lieberman, S, 1956, The elfects oi changes in roles on theattitudes of role occupants. Human Relations. 9: 385-402.

Lind, E. A., Lissak, R, I,, & Conlon, D, E, 1983. Decision controland process control eiiects on procedural fairness judg-ments./ourna/o/App/i'ed Socia/Psyciio/ogy, 13: 338-350.

Lind, E. A., & Tyler, T- R. 1988. The social psychology ofprocedural justice. New York: Plenum,

March, J. 1991. Exploration and exploitation in organization-al learning. Organization Science, 2: 71-87.

Mathieu, J. E., & Zajac, D. M. 1990. A review and meta-analysis of the antecedents, correlates, and conse-quences oi organizational commitment. PsychologicalBulletin, 108: 171-194.

McGregor, D. 1960. The human side of enterprise. New York:McGraw-Hill.

McPherson, J., & Smiih-Lovin, L, 1987. Homophily in voluntaryorganizations: Status distance and the composition oiface-to-iace groups. American Sociological Review, 52:370-379,

Meyer, G. W. 1994. Social information processing and socialnetworks: A test oi social influence mechanisms, HumanRelations, 47: 1013-1036.

Meyer, W., & Starke, E. 1982. Own ability in relation to seU-concepi of ability: A field study of information seeking.Personality and Social Psychology Bulletin, 8: 501-507.

Miceli, M. P., 8E Near, J. P. 1992, Blowing the whistle. NewYork: Lexington Books.

Miller, J. G. 1972. Living systems: The organization. Behav-ioral Science. 17: 1-82.

Morris, J. A., & Feldman, D. C. 1996, The dimensions, ante-cedents, and consequences of emotional labor. Acad-emy ol Management Review, 21: 986-1010.

Moskal, B. S. 1991. Is industry ready for adult relationships?Industry Week, January 21: 18-25.

Nemeth, C. J. 1985. Dissent, group process, and creativity.Advances in Group Processes, 2: 57-75.

Nemeth, C, J. 1997, Managing innovation: When less is more.California Management Review, 40(1): 59-74.

Nenieth, C, ]., & Wachter, J. 1983. Creative problem solving asa result oi majority versus minority influence. EuropeanJournal oi Social Psychology. 13: 45-55,

Parker, L. E. 1993, When to fix it and when to leave: Therelationship among perceived control, seli-eificacy, dis-sent and exit./ournai o/Applied Psychoiogfy, 78:949-959.

Pennebaker, J. W. 1985. Traumatic experiences and psycho-somatic disease: Exploring the roies of behavioural in-hibition, obsession, and confiding. Canadian Psychol-ogy, 26: 82-95.

Piefier, J. 1994. Competitive advantage through people. Bos-ton: Harvard Business School Press.

Pfeifer, J. 1997. New directions for organization theory. NewYork: Oxford University Press.

Pieifer, J,, & Leblebici, H. 1973. The effect of competition onsome dimensions of organizational structure. SocialForces, 52: 268-279,

Redding, W. C. 1985. Rocking boats, blowing whistles, andteaching speech communication. Communication Edu-cation. 34: 245-258.

Reichers, A, E,, Wanous, J. P., & Austin, J. T. 1997. Understand-ing and managing cynicism about organizationalchange. Academy ol Management Executive, 11(1):48-59,

Rice, R, E., & Aydin, C. 1991. Attitudes toward new organiza-tional technology: Network proximity as a mechanism{or social information processing. Administrative Sci-ence Quarterly. 36: 219-244.

Rosen, S,, & Tesser, A, 1970. On reluctance to communicateundesirable information: The MUM eifect. Sociometry.33: 253-263,

Rousseau, D., & Parks, J, M. 1993. The contracts of individualsand organizations. In L. L, Cummings & B. M. Staw (Eds.),Research in organizational behavior, vol. 15: 1-47.Greenwich, CT: lAI Press.

Rusbult, C. E., Farrell, D,, Rogers, G., & Mainous, A, G,, III.1988. Impact of exchange variables on exit, voice, loy-alty, and neglect: An integrative model of responses todeclining job satisfaction. Academy ol ManagementJournal, 31: 599-627.

Ryan, K. D., & Oestreich, D. K. 1991. Driving lear out oi theworkplace: How to overcome the invisible barriers (oquality, productivity, and innovation. San Francisco:Jossey-Bass.

Sachs, P. R. 1982. Avoidance oi diagnostic information inseli-evaluation of ability. Personality and Social Psy-chology Bulletin. 8: 242-246.

Salancik, G. R., & Pieifer, J. 1978. A social iniormation pro-cessing approach to job attitudes and task design. Ad-minisfrafive Science Quarterly, 23: 224-253.

Saunders, D. M., Sheppard, B. H., Knight, V,, & Roth, J. 1992,Employee voice to supervisors. Employee Responsibili-ties and Rights Journal, 5: 241-259.

Schneider, B, 1990. The climate ior service: Application oi theconstruct. In B. Schneider (Ed.), Organizational climateand culture: 383-412. San Francisco: Jossey-Bass,

Schneider, B., & Reichers, A. E. 1983. On Ihe etiology otclimates. Personnel Psychology, 36: 19-39.

Scott, S, G,, & Bruce, R. A, 1994, Determinants oi innovativebehavior: A path model oi innovation in the workplace.Academy ol Management Journal, 37: 580-607.

Scott, W. G., & Hart, D, K, 1979. Organizafional America.Boston: Houghton Mifilin.

Seligman, M, E. P. 1975. Helplessness. San Francisco: Free-man.

Sennett, R, 1998. The corrosion oi character: The personal


consequences of worJt in (he new capitalism. New York:Norton.

Shaw, M. E. 1981. Group dynamics: The psycho/ogy of smallgroup hehavioi. New York: McGraw-Hill.

Sprague, J., & Ruud, G. L. 1988. Boat-rocking in the high-technology culture. American Behavioral Scientist. 32:169-193.

Spreitzer, G. M. 1995. Psychological empowerment in theworkplace: Dimensions, measurement, and validation.Academy of Management Joarnal, 38: 1442-1465,

Staw, B. M., Sandelands, L. E., & Dutton, I. E. 1981. Threal-rigidity efiects in organizational behavior: A multilevelanalysis. Administrative Science Quarterly. 26: 501-524,

Swann, W, B., & Read, S. I. 1981, Acquiring self-knowledge:The search tor feedback that fits. Journal of Personalityand Social Psychology. 41: 1119-1128.

Tushman, M., & Romanelli, E. 1985. Organizational evolution:A metamorphosis model of convergence and reorienta-tion. In L. L. Cummings 8f B. M, Staw (Eds.), Besearch inorganizaliona] behavior, vol, 7: 171-222. Greenwich, CT:JAI Press,

Tyler, T. R.. Rasinski, K., & Spodick, N. 1985. The influence ofvoice on satisfaction with leaders: Exploring the mean-ing of process control./ourna/ oi Applied Social Psychol-ogy, 15: 700-724.

Vancouver, J. B., & Morrison, E. W. 1995. Feedback inquiry:The effect of source attributes and individual differ-

ences. Organizationai Behavior and Human DecisionProcesses. 62: 276-285.

Wagner, W. G., Pfeffer, J,, & O'Reilly, C, A., III. 1984. Organi-zational demography and turnover in top-managementgroups. Administrative Science Quarterly. 29: 74-92.

Weick. K. 1995. Sensemaking in organizations. ThousandOaks, CA: Sage.

Whitener, E. M., Brodt, S, E., Korsgaard, M. A., & Werner, J, M.1998. Managers as initiators o! trust: An exchange rela-tionship framework for understanding managerial trust-worthy behavior. Academy oi Management f^eview. 23:513-530.

Williamson, O. E. 1996, Economic organization: The case forcandor. Academy o/ Management Review, 21: 48-57.

Withey, M, I,, & Cooper, W, H, 1989. Predicting exit, voice,loyalty, and neglect. Administrative Science Quarterly.34: 521-539,

Woodman, R, W., & King, D. C. 1978, Organizational climate:Science or folklore? Academy oi Management Review. 3:816-826.

Wortman, C, B., & Brehm, J. W. 1975. Responses to uncontrol-lable outcomes: An integration of reactance theory andthe learned helplessness model. In L. Berkowitz (Ed.),Advances in experimental scciai psychology: 277-336.New York: Academic Press.

Zand, D. E. 1972, Trust and administrative problem solving.Administrative Science Quarterly. 17: 229-239.

Elizabeth Wolfe Morrison is an associate professor of management and organization-al behavior at the Leonard N. Stern School of Business. New York University. Shereceived her Ph.D. in organizational behavior from Northwestern University. Herresearch interests include employee initiative and proactivity, extrarole behavior,newcomer adjustment, and psychological contract violation,

Frances I. Milliken is an associate professor of management and organizationalbehavior at the Leonard N, Stern School of Business, New York University, Her re-search focuses on how individuals and teams perceive and interpret events andtrends in the organization and in its environment. She is also interested in the effectsof diversity on groups and organizations, as well as in the ways in which individuals'behavior is shaped by the organizational context within which they function.

organizational silence: a barrier to … classes/fall 07/org psy/cases... · organizational...

Documents