Failures in Upward

Communication in Organizations:
Three Possible Culprits^
KARLEIVE H. ROBERTS
CHARLES A. O'REILLY, n i
University of Caiifornia, Berkeiey
The impact of trust in the superior, perceived infiuence
of the superior, and mobility aspirations of subordinates
on upward communication behavior was examined. While
the importance of trust as a facilitator of open information
exchange was supported, this was less true of influence of
the superior and mobility aspirations.
In organizations, communication (roughly, the transference of information from sender to receiver and the meaning inferred from that information) from lower to higher members in the hierarchy is vital. The extent
to which accurate information is passed during superior-subordinate interactions has implications not only for the attitudes and satisfaction of the
sender and receiver but also for organizational decision-making, performance, etc. (3,10). Certainly examples of problems with information
exchange in upward transmission are easy to find. However, the identification of antecedents of poor information transmission upward is more
difficult than merely pointing out their existence.
Nevertheless, there is sufficient research examining upward communication in organizational hierarchies to suggest that three interpersonal factors
may be consistently related to aspects of communication. These three are:
(a) subordinate's trust in his superior (6, 8, 14, 15, 18, 25); (b) subordinate's perception of his superior's influence over his future ( 1 , 2, 11,
18, 22, 23); and (c) subordinate's mobility aspirations (11, 13, 18).
Karlene H. Roberts (Ph.D.University of California, Berkeley) is Assistant Professor of
Business Administration, University of California, Berkeley, California.
Charles A. O'Reilly, III, is a Ph.D. candidate. School of Business Administration, University of California, Berkeley, California.
iThis study was partially supported by Public Health Service Grant MH22054-01 and
Office of Naval Research Grant N00314-69-A-0200-1054.
205

206

Academy of Management Journal

June

In spite of the seemingly large amount of research, the evidence concerning the impact of these variables on upward communication is still
incomplete. First, many of the studies cited were conducted in laboratory
settings which may not be representative of the real world (24). Second,
different researchers approach the same variables from different perspectives. For example, in one study the concept of "influence" is operationalized as "status," in another as "prestige," and in still another as "power."
Similarly, "accuracy" in one case may be treated as "agreement" while in
another it refers to "openness." Differences in operationalizing variables
make cross-validation difficult. The end result is that while there exist
many studies which point to the import of trust, influence, and mobility as
potential moderators of upward communication behavior, speciflc conclusions are difficult to generalize.
The research reported here examines the relationship of the three interpersonal variables (trust, influence, and mobility) to a number of responses
to questions concerning communication with one's superior. These data
were gathered in four organizations and were intended to focus directly
on the real world aspects of superior-subordinate interaction and communication. Test-retest reliability and discriminant validity data for the
independent variables are included.
METHOD
Subjects were drawn from four diverse organizations. Organization A
is a state mental health outpatient facility with a team centered approach
to mental health care in which status differentials are deliberately minimized.
Completed questionnaires were collected from 101 members at all job levels,
a 98 percent response rate. Organization B is a high technology military
unit, in which 95 questionnaires were administered to both officers and
enlisted men, a 55 percent response rate. Organization C included 54
nursing and clerical employees in a medical center emergency room and
outpatient facility, a 100 percent response rate. Organization D was represented by 179 respondents from six branch offices of a flnancial institution
located in Great Britain and Ireland, a 90 percent response rate. All subjects were English speaking. In the organizations surveyed, all respondents
had a superior to whom they were required to pass information.
As a part of a larger questionnaire, each respondent provided answers
to seven questions assessing his trust in his immediate superior, the perceived influence of his superior, and his own mobility aspirations. These
items were pretested on a sample of 42 managers attending an evening
business course for test-retest reliability. To demonstrate partially the
discriminant validity of the items, a factor analysis was carried out using
the 95 subjects from organization B. The seven items, test-retest reliabilities
for the three indices, and the factor analytic results are reported in Table 1.

1974

Volume 17, Number 2

207

TABLE 1
Test-Retest Reliability and Factor Analytic Results for
Items Measuring Trust in Superior, Perceived Influence of the Superior,
and Mobility Aspirations of the Respondent

Items
Trust
1. How free do you feel to discuss with your
immediate superior the problems and difficulties in your job without jeopardizing your
position or having it held against you later?
2. Imrnediate superiors at times must make
decisions which seem to be against the interests of subordinates. When this happens
to you as a subordinate, how much trust do
you have that your immediate superior's
decision was justified by other considerations?
3. To what extent do you have trust and confidence in your immediate superior regarding his general fairness?
Influence
4. In general, how much do you feel that your
immediate superior can do to further your
career?
5. How much weight would your immediate
stfperior's recommendation have in any decision which would affect your standing, such
as promotions, transfers, etc.?
Mobility
6. As a part of your present job plans, do you
want a promotion to a higher position at
some point in the future?
7. How important is it for you to progress in
your present organization?
Proportion of communality accounted for:

Index
Test-Retest
Reliabilities
(N = 42)

.69

.62

.81

Varimax Factor
Loading
(N = 95)
1

.68

.22

.13

.76_

.08

.41

J9^

.23

.01

.28

.30

.81

.14

.16

.12

.90

.21

.23
.26

.88
.26

.20
.24

Items scored on seven point scales.

In addition to these items, 13 other questions were answered by each

respondent about various aspects of his upward communication behavior.
These questions centered on the relationship of the respondent to his
superior (time spent interacting, desire for interaction, and satisfaction with
communication) and on facets of his communication behavior with his
superior (modality use; amount of information sent and received; tendencies
to summarize, witiihold, or distort information; and feelings of information
overload). The items and their test-retest reliabilities are presented in
Table 2.
To examine the impact of the interpersonal variables on these 13
items, the three questions assessing trust and the two each tapping influence
and mobility were summed to form three indices. Correlations were then
computed for each organization relating each of the indices to the communication variables.

208

Academy of Management Journal

June

TABLE 2
Questions on Aspects of Upward Conununication Behavior
Test-Retest
Reliabilities
(N = 42)

Items ^
1. While working, what percentage of the time do you spend in contact
with your imniediate superior?
2. Of the total time you spend receiving inforaiation at work, what percentage comes from your immediate superior?
3. Of the total time you spend sending information at work, what percentage goes to your immediate superior?
Of the total time you engage in communications while on the job,
about what percentage of the time do you use the following methods
to communicate?
4. Written
5. Face-to-face
6. Telephone
7. Of the total amount of information you receive at work, how much
do you pass on to your immediate superior?
8. Are there forces leading to accurate or distorted upward information?
9. When transmitting information to your immediate superiors how
often do you summarize by emphasizing those aspects which are
important and minimize those aspects which are unimportant?
10. When receiving information from the sources listed below, how accurate would you estimate it usually isfrom your immediate superior?
11. Do you ever feel that you receive more information than you can
effectively use?
12. How desirable do you feel it is in your organization to be in contact
frequently with your immediate superior?
13. Put a check under the face that expresses how you feel about communication in general, including the amount of information you receive, contacts with your superiors and others, the accuracy of information available, etc.*"

.62
.82
.85

.56
.41
.58
.53
.53
.32
.52
.69
.64

.73

= Items scored on seven point scales.

*> Item used is a modification of the GM faces scale (25).

RESULTS
Trust and Upward Communication
The relationships of trust and upward communication for each of the
four organizations are presented in Table 3. The trust index is significantly
related in all four organizations to the subordinate's estimate of the accuracy
of information received from his superior and to his perception of his
superior's influence. This is a positive relationship showing that when a
subordinate expresses high trust in his immediate superior, he also believes
the information he receives from him is accurate and perceives his superior
to have high influence. In addition to these two variables, trust is significantly
related to desire for interaction and satisfaction with communication in
three of the four organizations. Again the relationship is positive, indicating
that high trust subordinates both desire interaction with their superiors and
are satisfied with communication in general. Two final relationships hold
across two of the four organizations. Low trust is associated in these cases
with the subordinate's disclosed tendency to block or withhold information

1974

Volume 17, Number 2

209

TABLE 3
Significant Correlations Between the Subordinate's Trust in Superiors
and Aspects of Upward Communication with the Superior
Organizations
Upward Communication Measures

A
(N = 101)

1. Perceived influence of the superior

2. Respondent's mobility aspirations
3. Percentage of time in contact
with superior
4. Percentage of information received from superior
5. Percentage of information sent to
superior
Percentage of time the following
methods are used to communicate:
6. Written
7. Face-to-face
8. TelephMie
9. Propensity to withhold information
10. Forces leading to distortion
11. Propensity to summarize information
12. Estimate of accuracy of information received from superior
13. Overload of information
14. Desire for interaction with superior
15. Satisfaction with communication
in general

.36***

B
(N = 95)
.25*
.37***

C
(N = 54)

D
(N = 179)

.28"

.34***

.63***
.43**

.32*
.24*

.34*

22*

.40***

25*
.50*

.39*

26*

.32*

.39***

36***

.56***

.38***

.43**

.41***

* p < .05

*p<.01
** p < .001

(gatekeeping) and his acknowledgement that there exist forces leading

to the distortion of upward information flow.
Influence and Upward Communication

Table 4 presents the intercorrelations for the influence index and upward
communication. Three relationships exist across the four organizations. A
subordinate's perception of his superior as having high influence is directly
related to high desire for interaction with the superior, high estimated
accuracy of information received from the superior, and high trust in the
superior. High perceived influence is also associated with high interaction
in three organizations, and with high satisfaction with communication in
two of the four organizations. As indicated by the lower correlations in
Table 4 compared to Table 3, influence does not seem to be as strongly
related to distortion or blockage of information as is trust.

210

June

Academy of Management Journal

TABLE 4
Significant Correlations Between the Perceived Influence of the Superior
and Aspects of Upward Communication witii the Superior
Organizations
Upward Communication Measures
1. Trust in the superior
2. Respondent's mobUity aspirations
3. Percentage of time in contact
with superior
4. Percentage of information received from superior
5. Percentage of information sent to
superior
Percentage of time the following
methods are used to communicate:
6. Written
7. Face-to-face
8. Telephone
9. Propensity to withhold information
10. Forces leading to distortion
11. Propensity to summarize information
12. Estimate of accuracy of information received from superior
13. Overload of information
14. Desire for interaction with superior
15. Satisfaction with communication
in general

A
(N = 101)
.36***
.27**

B
(N = 95)
.25*
.43***

C
(N = 54)
.28*
.31*

D
(N = 179)

.33*

.22**

.34***

.36*
.17*

.28*

.29***
.22**
21*
,44***

.27**

.34*

.23**

.21*

.29**

.46***

.32***

.33*

.16*

P < .05
p < .01
*** p < .001

Mobility Aspirations and Upward Communication

The relationship of respondents' mobility aspirations to aspects of
upward communication is shown in Table 5. Contrary to Read (18),
mobility is not strongly related to the accuracy of communication. It is
related in three of the samples to perceptions of high superior influence.
In two cases, mobility positively correlates with estimates of accuracy of
the information received by the subordinate from the superior and with
the subordinate's satisfaction with communication.
Trust, Influence, and Mobility
Of the three variables, trust and influence appear most closely related
to the respondents' estimates of their upward communication behavior. Responses to the questions about desire for interaction with superiors, estimates
of accuracy of information received from superiors, and satisfaction with
communication are largely attitudinal. Both trust and perceived influence
seem closely related to these. Questions about withholding information and

1974

Volume 17, Number 2

211

TABLE 5
Significant Correlations Between Subordinates' MobiUty Aspirations
and Aspects of Upward Communication with tiie Superior
Organizations
Upward Communication Measures
Perceived influence of the superior
Trust in the superior
Percentage of time in contact
with superior
Percentage of information received
e i e d from superior
inform
Percentage of information
sent to
superior
Percentage of time the following
methods are used to communicate:
6. Written
7. Face-to-face
8. Telephone
9. Propensity to withhold information
10. Forces leading to distortion
11. Propensity to summarize information
12. Estimate of accuracy of information received from superior
13. Overload of information
14. Desire for interaction with suerior
perior
15. Satisfaction with communication
in general

A
B
(N = 101) (N = 95)
.27**

.43***
.37***

C
(N = 54)

D
(N = 179)

.32*

.22*
.27**

.25*

.27**
.28*

.23*

.24*

*P<.05
** p < .01
***p<.001

distortion, while still attitudinal, are expressions of how the subordinate

may act. Trust in the superior appears most closely related to these variables.
No consistent relationships are evident among trust, influence, and
mobility and the other aspects of upward communication assessed. For
example, a subordinate's trust in his superior does not seem to alter the
percentage of information he transmits to the superior.
This corroborates a similar finding (16) based on three laboratory
experiments manipulating trust and influence. Using two data collections
from Organization A (seven months apart), and two collections from the
class of 42 managers (four weeks apart) cross-lagged panel correlations
were coniputed. The stabilities of the trust, influence, and mobility measures
were again shown. The cross-lagged correlations suggest that trust impacts
more on both gatekeeping and desire for interaction with the superior than
the opposite. Perceived influence of the superior and mobility aspirations of
the respondent did not have the same significant, well defined effects on
either gatekeeping or desire for interaction. Influence did, however, show
an antecedent relationship to satisfaction with communication.

212

Academy of Management Journal

June

DISCUSSION
In spite of its acknowledged importance, communication remains an
amorphous concept which is difficult to define and measure. Dance (4)
attested to this when he reviewed over 95 definitions of communication
and found them to represent 15 somewhat disparate conceptual themes.
Reviews of the literature (9, 17, 19, 20, 21) on organizational communication point out the difficulties encountered in investigating and measuring
facets of this variable.
In this research, the measures of communication behavior are weak
because of their perceptual nature and because they are admittedly an
incomplete description of the communication behavior possible in a
superior-subordinate relationship. However, one defense is that a subordinate's communication behavior is based on his perceptions of his interactions with his superior, not on objective measures of facets of the
interaction. Thus, a subordinate's processing of received information is
affected by his feelings about accuracy, not on an objective measure of
accuracy. To the extent this reasoning is correct, perceptual rather than
objective measures are justified.
The incompleteness with which the superior-subordinate exchange is
measured reflects, in part, the larger problem of definition and explication
of communication itself. A clearer understanding is needed of the importance of various aspects of communication which operate in different
circumstances. The aspects of communication measured in this study fall
into three general classes: questions designed to assess the likelihood that
information is distorted or withheld entirely, questions aimed at uncovering
attitudinal responses to aspects of communication, and questions designed
to gather data to describe how superiors and subordinates communicate
(for example, modality use, interaction frequencies, etc.). The presentation
of reliability data and the use of cross-validation are attempts to avoid
being misled through the use of what are otherwise relatively weak
measures.
As indicated previously, there is evidence that trust, influence, and
mobility are factors which have an impact on communications. The findings of this study, however, clarify and in several instances offer new insights into "old" variables.
Trust
Intuitively, interpersonal trust seems an important antecedent to the
openness and accuracy with which people, including superiors and subordinates, interact. Gibb (7), in a discussion of the climate for trust
formation, noted the deleterious effects produced by lack of trust on data
flow. Under conditions of low trust, groups have "inadequate data-flow,
which, in turn, causes groups to continue to operate with partial efficiency"
(7, p. 285). Gibb cites distortion of data through channels, grapevine
behavior, ambiguity, and protective phraseology as reactions to a lack of

1974

Volume 17, Number 2

213

a trusting climate. Friedlander (6), in a longitudinal study, supported

Gibb's contention that trust is a facilitator for group accomplishment. Zand
(25) also found that in high trust groups there was less socially generated
uncertainty and more efficient problem solving.
The findings reported here offer further empirical support for the importance of trust. Insofar as group accomplishment is related to the accurate
and open exchange of information, a lack of trust may impede data flow
and reduce performance. Although performance data were not available
for this study, the expression of low trust in superiors is consistently related
to: (a) those variables which may signify potential gatekeeping and distortion of information (acknowledgement of the presence of influences
leading to distortion of upward information flow, propensity to withhold
information, and lack of desire for interaction); and (b) attitudinal variables (doubts about the accuracy of information received and dissatisfaction with communication in general). In these circumstances the untrusting
subordinate has little desire for interaction with his superior. Certainly in
an ambience like this it is easy to envision groups operating with the inadequate data flow and partial efficiency proposed by Gibb (7), Friedlander (6), Mellinger (15), and others.
Influence
The impact of the superior's influence over the subordinate on upward
communication is not as clearly defined as that of trust. Intuitively, the
idea that a superior perceived as having high influence might be one to
whom communication is guarded seems reasonable. Table 4 shows that
subordinates with high influence superiors desire interaction and believe
that accurate information is received from them. Interestingly, however,
influence does not exhibit any consistent relationship with the variables
which might be precursors to inaccurate communications, i.e., distorting
pressures or tendencies to withhold information. It is related, however, to
trust, mobility, and satisfaction with communication in general.
This picture differs slightly from those portrayed in other investigations.
For example, Kelley, in a laboratory experiment, concluded in part that
"the existence of a hierarchy produces restraining forces against communicating criticisms of persons at the other level" (11, p. 56). Festinger
(5) also pointed out that hierarchies restrain free communication, particularly criticism and aggressively-toned comments by low status members
to high status ones. This might result in tendencies on the part of subordinates to withhold information from the superior which could be critical to
the latter's job. Yet, the results of this study strongly suggest that the effects
of influence may not be independent of those of trust. A laboratory investigation of the impact of trust and influence on information filtration
corroborated the predominance of trust over influence (19).

214

Academy of Management Journal

June

Mobility
The picture of a highly ambitious and upwardly mobile organizational
member may easily include the idea of a person who restricts his communication to exclude unfavorable or critical information. In pursuit of
his own advancement, such a person might alter his communication behavior to achieve his desired goals. Read (18) offered this mechanism as
his major hypothesis, predicting a negative relationship between upwardly
mobile members of an industrial organization and the accuracy of their
upward communication. He confirmed this hypothesis and concluded that
high mobility aspirations strongly militate against the accurate communication of potentially threatening information even when high trust prevailed.
The results reported here question this conclusion. The lack of strong
correlations between mobility aspirations and communication behavior
suggests that the impact of mobility may operate only in certain groups.
Recall that the subjects sampled in this study are drawn from a variety of
work settings and positions in the organizational hierarchy. Read focused
on a set of 52 executive pairs. Perhaps with highly mobile managers communication may be altered more. Even though the sample of 42 managers
attending an evening MBA program might be reasonably labelled as having
high mobility aspirations, no important relationships were discovered between mobility aspirations and the other measures discussed here. Only
in Organization B, the military sample, did upward mobility aspirations
seem strongly related to upward communication behavior. For the larger
picture, however, the impact of mobility was not great.
While the nature of the data presented does not permit strong inferences
to be made about the relative importance of the interpersonal variables,
a trend is apparent; trust in one's superior appears related to communication
openness. Both trust and influence are related to attitudes about superiorsubordinate interaction. Mobility aspirations seem less important a factor
than reported previously. Certainly the results reported here may be interpreted as offering some empirical support for the proposition that trust
is a facilitator for open and accurate upward communication.
REFERENCES
1. Alkire, A., M. Collum, J. Kaswan, and L. Love. "Information Exchange and Accuracy
of Verbal Behavior Under Social Power Conditions," Journal of Personality and Social
Psychology, Vol. 9 (1968), 301-308.
2. Cohen, A. "Upward Communication in Experimentally Created Hierarchies," Sociological Review, Vol. 15 (1958), 342-351.
3. Collins, B., and B. Raven. "Group Structure: Attraction, Coalitions, Communication
and Power," in G. Lindzey and E. Aronsen (Eds.), The Handbook of Social Psychology,
Vol. 4 (Reading, Mass.: Addison-Wesley, 1968), pp. 102-104.
4. Dance, F. E. X. "The 'Concept' of Communication," Journal of Communication, Vol.
20(1970), 201-210.
5. Festinger, L. "Informal Social Communication," Psychological Review, Vol. 57 (1950),
217-282.
6. Friedlander, F. "The Primacy of Trust as a Facilitator of Further Group Accomplishment," Journal of Applied Behavioral Science, Vol. 6 (1970), 387-400.

1974

Volume 17, Number 2

215

7. Gibb, J. R. "Climate for Trust Formation," in L. P. Bradford, J. R. Gibb, and K. D.

Benne (Eds.), T-Group Theory and Laboratory Method: Innovation in Re-Education
(New York: Wiley, 1964), pp. 279-309.
8. Giffin, K. "The Contribution of Studies of Source Credibility to a Theory of Interpersonal Trust in the Communication Process," Psychological Bulletin, Vol. 68 (1967),
104-120.
9. Guetzkow, H. "Communications in Organizations," in J. March (Ed.), Handbook of
Organizations (Chicago: Rand-McNally, 1965), pp. 534-573.
10. Indik, B., B. Georgopoulos, and S. Seashore. "Superior-Subordinate Relations and Job
Performance," Personnel Psychology, Vol. 14 (1961), 357-374.
11. Kelley, H. "Communication in Experimentally Created Hierarchies," Human Relations, Vol. 4 (1951), 39-56.
12. Kunin, T. "The Construction of a New Type of Attitude Measure," Personnel Psychology, Vol. 8 (1955), 65-78.
13. Lawler, E., L. Porter, and A. Tennenbaum. "Managers' Attitudes Toward Interaction
Episodes," Journal of Applied Psychology, Vol. 52 (1968), 432-439.
14. Loomis, J. L. "Communication, the Development of Trust, and Cooperative Behavior,"
Human Relations, Vol. 12 (1959), 305-315.
15. Mellinger, G. "Interpersonal Trust as a Factor in Communication," Journal of Abnormal and Social Psychology, Vol. 52 (1956), 304-309.
16. O'Reilly, C , and K. Roberts. "Information Filtration in Organizations: These Experiments," Organizational Behavior and Human Performance (in press).
17. Porter, L. W., and K. Roberts, "Organizational Communication," in M. Dunnette (Ed.),
Handbook of Industrial and Organizational Psychology (Chicago: Rand-McNally, in
press).
18. Read, W. "Upward Communication in Industrial Hierarchies," Human Relations, Vol.
15 (1962), 3-16.
19. Roberts, K., and C. O'Reilly. "Some Problems in Measuring Organizational Communication, Journal of Applied Psychology (in press),
20. Roberts, K., C. O'Reilly, G. Bretton, and L. Porter. "Organizational Theory and Organizational Communications: A Communication Failure?" Human Relations (in press).
21. Thayer, L, "Communication and Organization Theory," in F. E. X. Dance (Ed),,
Human Communication Theory: Original Essays (New York: Holt, 1967).
22. Wager, L, "Organizational 'Linking Pins,' Hierarchical Status and Communicative
Roles in Interlevel Conferences," Human Relations, Vol. 25 (1972), 307-326.
23. Watson, D. "Effects of Certain Social Power Structures on Communication in Taskoriented Groups, Sociometry, Vol. 28 (1965), 322-332.
24. Wieck, K. "Laboratory Experimentation with Organizations," in J. March (Ed.), Handbook of Organizations (Chicago: Rand-McNally, 1965), pp, 194-260.
25. Zand, D. E. 'Trust and Managerial Problem Solving," Administrative Science Quarterly, Vol. 17 (1972), 229-240.