Upward communication filtering in the project management environment


Stephen E. Barndt

Pacific Lutheran University

Tacoma, Washington

The communication of information is essential to cooperative goal directed behavior. Formal organizations rely on a free flow of information up and down the scalar chain. However, free flow does not mean an open pipeline that transfers all information. Rather, organization members are expected to act as filters —passing on to supervisors and subordinate only that information needed for efficient performance of duties and satisfaction of personal needs. The extent to which supervisors and subordinates, alike, filter out the unnecessary information and transmit the important information should be expected to be a factor impacting organizational performance. An obvious managerial function is to encourage narrowing down the volume of information transmitted and, at the same time, lessen the exclusion of important information.

Studies have not only confirmed the intuitive belief that less than the total of all information is passed along but have provided some quantitative measures as well. For instance, Boyles and Wicker (1977), using a behavioral laboratory experiment involving subjects acting out roles in research and development projects, found that only approximately 54 to 65 percent of available information was passed upward from subordinates to supervisors. Data presented by O‘Reilly and Roberts (1974) indicated that subjects role playing managers in a staff specialist department (personnel), communicated, on the average, approximately 73 percent of available information. Thus, allowing for differences in the kinds of settings, tasks, and information available, one should be prepared to only see in the neighborhood of 55 to 75 percent of the problem-related information potentially transmittable to immediate supervisors actually transmitted.

The filtering of from 25 to 45 percent of information in upward communication is not, in itself, a matter of concern. In fact, if the information filtered out is unneeded and if no important information is excluded, the communication process is working just as we’d like. Unfortunately, some important information is likely to be withheld for one or more of a variety of conscious or nonconscious reasons. Both the Boyles and Wicker (1977) and O‘Reilly and Roberts (1974) studies provide indications of what might be expected in the magnitude of such filtering of important information (see table 1). In both studies, independent judges pre-identified important items of information as such. Using the percent of the total important information available that was actually transmitted upward as a measure, the two studies yielded similar results. In the O‘Reilly and Roberts research, only 72 percent of available important information was transmitted while in the Boyles and Wicker research only 63 percent was transmitted in one experimental setting and only 77 percent in another. Thus, to the extent that the laboratory experiments reflect the filtering that exists in the work setting, one could expect that supervisors may be denied one fourth or more of the information that they should have to perform their jobs.

An additional indication of the faialure to accurately and usefully communicate to superiors is provided by Boyles and Wicker. Their data show that only 83 to 85 percent of the total information passed upward was classified as important. This means that, in addition to denying supervisors some of the information they need, subordinates also send along unimportant information.

Reasons for Filtering

In an era before much research was conducted on organizational communication, Simon (1957) provided his insights concerning several conditions each of which will result in the upward communication of information. First, he indicated that its transmission must not result in consequences unpleasant for the sender. Second, if the superior would obtain the information anyway, the subordinate is likely to go ahead and tell him first. Third, if the superior needs the information to deal with his superiors, would be embarrassed without it, the subordinate is likely to attempt to avoid displeasure by passing the information along. However, starting at about the time of Simon’s statements, research interest has been kindled in a number of facets of communication. One result is that a considerable, although incomplete, research based body of knowledge now exists concerning blocking, withholding, or otherwise passing to superiors less than the total amount of information possessed by a subordinate. This research has explored a number of variables posited as being related to content and accuracy of upward communication. These variables include mobility, status, ascendency drive, level of insecurity, trust, influence, autonomy, and organizational climate.

With respect to mobility opportunity and status, Kelley (1951) found that the task relevance of upward communication was related to both the communicator’s possibility for upward mobility and status position. That is, the more unpleasant the position, e.g., low status and nonmobile, the greater the tendency to transmit task irrelevant information. Similarly, Cohen (1958) concluded that content of communications in terms of length and amount of irrelevant information was related to the rank (based on power and status) and upward mobility of the communicator. In particular, low rank, nonmobile individuals tended to communicate more irrelevant information.

When upward mobility aspirations was tested as an explanatory variable, the conclusions become mixed. For example, Read (1962) found support for the existence of an inverse relationship between upward mobility and accuracy of upward communication. In addition, Athanassaiades (1973) found, with respect to two occupation groups, that distortion of upward communication was positively related to ascendency drive (similar to mobility aspiration) and level of insecurity. Further, O‘Reilly (1978) found indications that less intentional distortion was associated with low mobility aspirations. O‘Reilly’s conclusion, based on weak statistical support for this relationship, was that mobility aspirations are of little importance in explaining intentional distortion. This conclusion was also made by Burchette, Porter, Blalack, and Davis (1977) as a result of their study of communication in four organizations.

Support for the existence of relationships between both accuracy and openness of upward communication and trust in one’s superior and the influence of the superior has been provided by Read (1962); O‘Reilly and Roberts (1974); Burchette, Porter, Blalack, and Davis (1977); and O‘Reilly (1978). The degree of trust was found to be strongly and directly associated with the degree of open, accurate, complete, or undistorted communication in all studies. Influence of the superior, although directly correlated with open or undistorted communication, was considered to be of much lesser importance by O‘Reilly and Burchette, et al.

Finally, autonomy and organizational climate have been investigated as factors impacting upward communication. Autonomy of authority in performing work was found to be inversely related to the distortion of upward communication (Athanassiades, 1973). Recently Muchinsky (1977) presented weak support for the belief that accuracy of communication is positively related to several dimensions of organizational climate. More importantly, Muchinsky found a direct correlation between organizational climate and trust. This adds support to the belief advanced by communication theorists that a supportive climate builds interpersonal trust. Based on the research results presented above we can carry the causal inference further to suggest that trust leads to more open, more complete, less distorted communication.

Generally the research on communication supports the possibility of the creation of inaccurate, irrelevant, or even untrue information in the upward communication process. Further, it provides a basis for classifying the various behaviorally and structurally derived variables in terms of their probable impact on filtering. The results of such a classification are presented in table 2. As shown, trust and organizational climate are included in the highest potential category. This is because of a combination of strong relationships found through research and the opportunity of managers to, personally, do something about them. The inability of the individual manager to do much about the subordinate’s ascendency drive; his own (the supervisor’s) degree of perceived influence; or the structural aspects of the status attached to a job, the career mobility attached to a job specialty, or the security attached to a job are the principle reasons for categorizing these variables as low potential. Although the feeling of security and autonomy of work performance were only supported as upward communication influences in one study, creating a non-defensive, secure environment and providing limited opportunities for self directions are within the capability/authority of many managers. For this reason, a climate of autonomy and security are classified as variables of moderate potential for lessening filtering.

The Project Manager and Upward Filtering

The project manager’s role as a communicator is, on the one hand, particularly crucial and, on the other, particularly demanding. His failure to obtain information about problems, potential problems, and alternatives in time for effective redesign, rescheduling, or reallocation decisions can have adverse consequences on attainment of project tasks and the project as a whole depends on timely, pertinent information. However, the dynamic nature of the project means that needed information cannot always be determined beforehand. Therefore, in most projects, we find that standing formal reports, although important, supply but a fraction of the information actually needed by the project managers for decision making. This means that much of the needed information must come to the project manager from the people who have it —in nonspecific formats, by nonspecific channels, and at nonspecefic times.

The filtering of information in upward communication is highly probable in the project environment. There are three conditions that point to this high probability. First, project personnel don’t always know what is important or unimportant. The lack of clear cut channels of communication and lack of reporting formats can leave them uncertain of just what is wanted and when. In addition, team members typically lack knowledge of the total project and thus lack knowledge of what may be newly important at any given time. Second, in the absence of formal channels and reporting formats, the channels that are characteristically used may inadvertently cause filtering. In unrehearsed face-to-face or telephone conversation, it is easy to forget or to be led to a different topic before the last topic should have been concluded. Also, signs of impatience or tiring on the part of supervisors may cause subordinates to reduce message contact in order to “finish.” Last, project personnel may fear that they will be penalized for passing on unpleasant news or information they perceive will reflect unfavorably upon themselves. The changing composition of the project team and specialized and, sometimes even, isolated nature of their work means that many may not get to know their supervisor well enough to know if they can trust him with such information. The extent to which project supervisors can improve on the three conditions that otherwise favor filtering in the project organization will be reflected in an improvement in full disclosure of important information

Ways to Reduce Upward Filtering

The key aspects of any effort to effectively reduce upward filtering of important information are recognition, creation of a supportive climate, and creation of channels of communication.

The most crucial step toward reducing filtering must be a recognition that important information is filtered out in the upward communication process and a desire to improve on the situation. The figures given earlier in this paper, indicating that 23 to 37 percent of important information is filtered, provide a useful baseline for deciding whether or not to even bother. If these figures are acceptable, then the project manager can probably dismiss filtering as a problem area. If they are not acceptable, the project manager needs to (1) internalize the question “am I getting it all” when receiving a communication and (2) effect changes to improve the chance that he is “getting it all.”

Actual efforts should focus on those variables classified as high and moderate potential for manipulation to lessen filtering (table 2). Actually, supportive organizational climate and trust between supervisors and subordinates throughout the project organization are the only variables that need be directly addressed. In project organizations, team members typically already have a reasonably high degree of autonomy/freedom of action, and trust, itself, will lessen perceived threat against an individual’s security.

Since a supportive organizational climate and trust seem to be closely related, we will cover actions that can be taken by the project manager, in the normal course of his duties, to improve both at the same time. First, the project manager (and other project supervisors) needs to develop an effective presence. The project supervisor needs to be seen in a friendly, business-like manner and to project an image of professional competence. The finding by Burchette, Porter, Blalack, and Davis (1977) that trust in the superior is positively correlated with percent of information received from the superior, the percent of time in contact with him, and the perceived accuracy of information received from him, also indicates supervisors should devote time and attention to downward communication — as part of their job. To the extent that supervisors are willing to communicate their feelings, concerns, problems, and hopes to subordinates, such downward disclosure can be expected to serve as a model for the subordinate’s upward disclosure. Of course, the supervisor must be careful to not assume an undue role of confessor and counselor. Second, the supervisor should visibly demonstrate a commitment to the job. Again, this means assuming a role as a model to project team members. These team members may become committed to the project even though they may be there only for a relatively short time and may have no personal commitment to the project manager. A personal commitment to the project and hence a personal desire to not see it get in trouble can provide an incentive to more complete upward disclosure. Third, the supervisor must take time to listen and not, at the same time, appear as though he really does not want to listen or hasn’t 41 the time to listen to everything the subordinate has to say. He should accept what the subordinate has to say not only for its importance to himself but also for its importance to the subordinate. Finally, the supervisor needs to consistently reward open upward communication through acknowledgement, praise, and public credit.

As pointed out by O‘Reilly (1978), even with trust, there is a bias against passing unfavorable (though important) information upward. Therefore, the project manager needs to work further to provide a structure more conducive to upward communication. This structuring does not necessarily mean setting up rigid rules, procedures, offices, and otherwise attempting to prescribe exactly what will be communicated, when, and how. Rather, the approach taken here is to suggest those less formal kinds of structural adaptations that project managers and supervisors can easily make on their own. The first such adaptation is for the project manager to assume a major role as a communicator. This means he must plan on spending a lot of time communicating. He needs to establish himself as a communication focal point and prove out by seeking information, asking questions in a nonthreatening, nonevaluative manner, and passing on information to his subordinates. If the project is large, he may not be able to act as the total communications focal point. In this case, he will need to identify those persons or positions that seem to be communication nodes and establish more frequent and open communications with them. If no such communication liaison positions exist and if communication problems so warrant, capable communicators may be shifted so as to place them where they are most likely to receive and pass messages. A last structural change involves formally and informally establishing the general and specific kinds of information that are considered to be important. This does not mean creating a catalog of new standard reports. Standard reports are often not very applicable even in standard situations let alone project type non-standard situations. Therefore, it is probably better to not require such reports that waste subordinates’ time and cause them to question the supervisor’s competence. A more appropriate technique is to simply disseminate updated ideas about what is important in meetings, person to person conversation, and memoranda/project bulletins on a recurring, as needed basis.

In conclusion, lessening the extent of upward filtering of important information can be done by project managers without significant added cost and without special behavioral training. All that is needed is a commitment to being a better communicator within the context of the project manager’s job as presently defined. This commitment involves the project manager:

1. recognizing that upward filtering does occur, that it is extensive, and wanting to lessen it;

2. thinking of communication as a major supervisory role and spending a lot of time on it;

3. being a model of commitment to the job and to team members;

4. letting project team personnel know the kinds of information that are important;

5. becoming a communication focal point, identifying other focal point persons, and communicating with them;

6. actively interacting with and communicating information to subordinates;

7. actively seeking information, asking questions, and listening; and

8. rewarding upward communications.


1. Athanassaides, John C. “The Distortion of Upward Communication in Hierarchical Organizations.” Academy of Management Journal, Vol. 16 (1973), 207-226.

2. Boyles, David J. and Harvey L. Wicker. “Deliberate Filtering of Upward Communication and Awareness of Self and Others in Air Force Subordinate-Supervisor Dyads.” Unpublished Masters Thesis, number LSSR 26-77B, Air Force Institute of Technology, 1977.

3. Burchette, Hugh T., and Robert O. Porter, Richard O. Blalack, and Herbert J. Davis. “Gatekeeping and Upward Communication: Another Test of Three Contributing Factors,” paper presented at the 37th Annual Meeting of the Academy of Management, Orlando, Florida, August 1977.

4. Cohen, Arthur R. “Upward Communication in Experimentally Created Hierarchies,” Human Relations, Vol. XI (1958), 41-53.

5. Kelley, Harold H., “Communication in Experimentally Created Hierarchies,” Human Relations, Vol. IV (1951), 39-56.

6. Muchinsky, Paul M. “The Interrelationships of Organizational Communication Climate,” Proceedings of the 37th Annual Meeting of the Academy of Management, Orlando, Florida, August 1977, 371-374.

7. O‘Reilly, Charles A. III, “The Intentional Distortion of Information in Organizational Communication: A Laboratory and Field Investigation,” Human Relations, Vol. 31, No. 2 (1978), 173-193.

8. O‘Reilly, Charles A. III and Karlene H. Roberts, “Information Filtration in Organizations: Three Experiments,” Organizational Behavior and Human Performance, Vol. 11 (1974), 253-265.

9. Read, William H. “Upward Communication in Industrial Hierarchies,” Human Relations, Vol. 15 (1962), 3-15.

10. Simon, Herbert A. Administrative Behavior. 2nd. Ed. New York: The Free Press, 1957.


Comparison Percent Information Send to Supervisor
Boyles and Wicker Study O‘Reilly and Roberts Study
1st Experiment 2nd Experiment
Total information passed as percent of total information possible 54 65 73
Important information passed as percent of important information possible 63 77 72
Important information passed as percent of total information passed 83 85


Variable Potential
High Moderate Low
Ascendancy Drive X
Influence X
Security X
Trust X
Mobility X
Status X
Organizational X
Autonomy X
Security X



Related Content