Groupthink, Self Serving Subordinates and Uncertainty - Can They Be Overcome

In this article, Martin Evans considers the barriers to accurate information gathering in large organizations.

“I expect to get valid information … I can’t make good decisions unless I get valid information” George W. Bush, April 13^th 2004.

George Bush’s cry is echoed by every organizational manager in the world. Looking back on the Presidency of George W. Bush we can see many information failures. Every manager would like to be sure that the information received was both timely and accurate. Every good manager knows it’s their responsibility to make sure that information received is timely and accurate. Despite his Yale education, George W. Bush did not learn this. Every good manager knows about the three major barriers to the realization of good information: group think, self-serving subordinates, and uncertainty absorption. Despite his Harvard Business School Education, George W. Bush failed to grasp this. Good managers are proactive to ensure that these barriers are overcome. Despite his years of management experience, George W. Bush did not learn this.

In this article I will deal with the three barriers: first Groupthink, then self serving subordinates, and finally the most subtle of all, uncertainty absorption. I will describe each phenomenon and discuss ways of overcoming them.

Groupthink

The Bay of Pigs, the decision to bomb North Vietnam, and our adventure in Iraq provide us with ample illustrations of Groupthink. This magnificently Orwellian term was coined by psychologist Irving Janis to refer to a group’s inability to tolerate dissent, to a situation in which getting agreement around a solution becomes more important than developing the best solution.

“We have all experienced what the French call “the bottom of the stairs” feeling when one remembers that critical comment that might have changed the course of a discussion. This “last chance” meeting gives us the opportunity to make that comment. It is like those Viking chiefs who used to make their decisions twice over: once when drunk and once when sober. If the two agreed they would carry out the decision; if they differed they would think again.”

Effective managers know that tough action must be taken in order to prevent the domination of the forces of Groupthink. What does Groupthink look like? How would we recognize it in our own decision-making groups? Janis and earlier Norman Maier have identified a number of symptoms that we should be on guard against.

It is often difficult, because of the Groupthink phenomenon, for us to recognize when these symptoms occur in our own groups – after all we know better, we would never make the same mistakes. That illustrates the first and greatest problem: the group believes that it is right; that it is right both factually and morally, and consequently the scenario of the operation (Bay of Pigs, Iraq) will unfold as planned. This sense of invulnerability and moral correctness leads to four things that affect the group as it engages in the process of making a decision. First, there exists a sense of unanimity within the group; everyone agrees on a plan. Indications of this are a failure to generate more than one or two alternative courses of action, and an individual’s unwillingness to express his or her reservations – individuals are their own self-censors. They suppress and keep to themselves their doubts and reservations about the plan. Often, many people share these reservations and the group lacks a true unanimity. In addition, members of the group put subtle pressure on those whose questions slip through the guard of their own self-censoring. As Janis observes they do this by limiting the bounds of criticism to details of the plan rather than to its underlying assumptions and by isolating the dissenter into a slightly ridiculous role as in Lyndon Johnson’s term for Bill Meyers: “Let’s hear what Mr-Stop-the-Bombing has to say”. Finally and most injuriously the group develops the MindGuards. These are the people who filter the information coming to the group. They make sure that outside information is suppressed or reinterpreted if it fails to support the cherished assumptions of the group; a good example was the sidelining of General Shinseki when he suggested that several hundred thousand troops would be required to pacify post-war Iraq. As a result of this process, the group makes its decision only upon information that is supportive of that decision. This builds up a self-fulfilling cycle of correctness. The illusion of rightness and unanimity is preserved; no disruptive questioning or information is admitted by the group.

These processes are pervasive in decision-making groups. One of the few ways of reducing their severity is to institutionalize dissent. Janis and Maier prescribe several mechanisms for doing this. They represent a re-establishment of the traditional American system of checks and balances in the political process.

“Messages are transmitted up organizational hierarchies and each level has its own concerns: folk at the bottom are concerned with operational nitty-gritty issues; those in the middle worry about administrative issues; those at the top are focused on strategic issues.”

A group should routinely make decisions in a two stage process. First, critical evaluation should be suspended and a wide variety of alternative courses of action generated; the final list should consist of the off-beat as well as the obvious. This can be done by having several subgroups or individuals brainstorm lists of alternatives which are then brought back to the policy making group. In addition, group members should be encouraged to discuss the issues with their fellow workers and subordinates so as to bring a wider range of alternatives to the group’s attention.

Second, each alternative should be criticized in terms of its strengths and weaknesses, with equal time given to both aspects. Work should be done on integrating several flawed solutions into a better one. This can be done by the leader of the group formally assigning the responsibilities of criticizing the alternatives to each group member. This alone is not enough; the leader must accept with good grace and an open mind the criticisms made about his ideas and proposals. Without this, all criticism will degenerate to a few pro forma comments. It is also helpful to augment the group, from time to time, with outside members who can bring a wider range of criticisms to bear upon the alternatives under consideration. These criticisms are apt to be less inhibited than those of the group members as they will have not built up any ownership of the alternatives under discussion.

Third, when the group is close to reaching a decision about the best alternative, the leader should appoint a Devil’s Advocate for the two or three options that are under serious consideration. These individuals are to challenge the assumptions and expectations of the proponents of each alternative.

Finally, as a group approaches a final decision, it should be augmented by outsiders who have not been involved in the decision so far. These people will be able to bring to bear a wider variety of perspectives and their comments and criticisms will be less inhibited than those of group members as they will not have built up any ownership of the proposed solution.

At the last step – once a decision has been taken – the group should meet one more time to review doubts and challenge the correctness of the decision. We have all experienced what the French call “the bottom of the stairs” feeling when one remembers that critical comment that might have changed the course of a discussion. This “last chance” meeting gives us the opportunity to make that comment. It is like those Viking chiefs who used to make their decisions twice over: once when drunk and once when sober. If the two agreed they would carry out the decision; if they differed they would think again.

These practices will not completely eliminate errors of judgement, nor will they guarantee success as even the best laid plans may go awry. These procedures will however minimize the chance of failure due to overlooking possible alternatives or failing to consider the potential side-effects of the outcome selected.

Of course, the successful implementation of each of these safeguards depends critically upon the role of the group’s leader. The leader must show that he/she values dissent and should provide a strong role model of the acceptance of critical comment. This means that the group leader should keep an open mind, be open to criticism, and not commit her or himself to a course of action until all the alternatives have been thoroughly explored.

“R&D people are concerned with technical and, to a lesser extent, business issues; Marketing people were more balanced having almost equal concern with business issues, customer needs, selling and technical issues; However, managers in Manufacturing were mainly concerned with production issues and, to a lesser extent, technical. These different foci make it difficult to share insights.”

Self serving subordinates

In organizations, information is power. By holding back information, an individual makes others dependent on him in their decision making processes. In other situations, where there is a major difference in the technical competence between a manager and a subordinate (so the boss is unaware that the wool is being pulled over his eyes, the subordinate can manipulate the flow of information so as to make sure the boss makes decisions that are consistent with the subordinate’s best interests even though they may be sub-optimal for the organization. The classic description of this process was provided by Andrew Pettigrew many years ago in his  graphic description of the power of an information gatekeeper in affecting the outcome of a strategic organizational decision (to make a major computer purchase). By carefully controlling the flow of information between a set of suppliers and the decision  making body (the Company’s Board of Directors), the gatekeeper ensured that ‘his’ supplier received the contract for the purchase.  The gatekeeper used the following tactics:

a) Providing fulsome and timely replies to communications from the supplier that he favored, but only reluctant and tardy replies to communications from other suppliers.

b) Refusing to visit or be “wined and dined” by the other suppliers.

c) Refusing to let his subordinate managers (each of whom  favored a different supplier based upon their departmental interests; this was after all a multi-party political game) have direct access to the deliberations of the decision making body.  The only exception to this was on an occasion when he knew that the most influential board member would be absent.

d) Providing a biased balance of positive and negative information about ‘his’ supplier and the other suppliers to the decision making body.  He did this by diagnosing the ‘assessed stature’ (how good it looked in the eyes of the Board) of his department with the decision  making body; he then passed information that he wanted to be believed when his stature was high (positive information about ‘his’ supplier, negative information about other suppliers), and he passed information that he wanted to be ignored or discounted (negative information about ‘his’ supplier, positive information about the others) when his department’s stature was low. Note that over the whole decision period he provided balanced positive and negative information about each supplier. The imbalance lay in his strategic shifting of the balance of positive and negative information depending on how his bosses perceived him.

How can managers prevent being manipulated in this way? The first shield against this kind of manipulation is by being as technically well informed as the subordinate. This enables the manager to undertake his own evaluation of the adequacy of the information. However, this shield is rarely available in this world where problems are multifaceted and it is impossible for the manager to be an expert in each facet. The second, more proactive, technique is for the boss to reach down the hierarchy to gain information directly from the subordinate’s own subordinates (and the suppliers) rather than having it filtered through the bottleneck of the subordinate’s filtering mechanisms.

Further distortion by self-aggrandizing subordinates is the suppression of bad news or the inflation of good news. This tendency is increased when several subordinates are in competition for the support of the boss and when, as in this case, the boss has the power to decide the fate of those subordinates (in terms of budgetary outcomes). This is what appears to have happened in the famous August 6, 2001 Presidential Daily Briefing document. The FBI claimed that they were engaged in seventy “full field” investigations of terror suspects. In fact they were merely passively monitoring the suspects’ financial affairs rather than their quotidian activities (like learning to fly). Again, a vigilant manager would look beyond the label “full field” and ask what exactly was being done, to whom, and why, and where. Tough questioning might have revealed how sham these “full field” investigations were with a consequent increase in surveillance of the suspects. Other things that can be done to minimize this tendency to inflate good news and suppress bad news is through the development of trust between the parties – the boss and each of the subordinates with each other.

Uncertainty Absorption

The third prevalent source of information distortion in organizational communication is uncertainty absorption. This occurs where raw data are summarized, aggregated, and edited prior to being transmitted onward.  Two processes seem to be involved: the selection of data from the plethora of incoming stimuli; and the packaging of this data for transmission.  In both processes, the monitor/communicator’s frame of reference or cognitive map is crucial in determining what information will be noticed and what information will be transmitted. In either process, information can either be suppressed, attenuated or enhanced.

It is clear that organizational position (functional specialty, staff/line, hierarchical level) affects the frame of reference that individuals use to scan their environments and incoming communication messages. These differences are found both in terms of the aspects of the environment attended to and in terms of the complexity of the individual’s cognitive map. For example, higher level managers in both staff and line functions had higher cognitive complexity than those managers at lower levels; however, high level staff managers were less complex (more single tracked) than their line counterparts.

The problem is that information congruent with a particular frame of reference is more easily noted or transmitted.  When people with different frames of reference communicate, the receiver will distort the incoming information to fit her/his frame of reference.  In organizations,  different departments have different “funds of knowledge” and different “frames of reference.” For example, R&D people are concerned with technical and, to a lesser extent, business issues; Marketing people were more balanced having almost equal concern with business issues, customer needs, selling and technical issues; However, managers in Manufacturing were mainly concerned with production issues and, to a lesser extent, technical. These different foci make it difficult to share insights. Deborah Dougherty found that successful product development only occurred when firms broke out of these habituated ways of thinking. Successful innovation resulted in someone, somehow ensuring that multiple frames of reference were considered. Only when issues and solutions were considered in these varieties of ways was success ensured.

Similarly, messages are transmitted up organizational hierarchies and each level has its own concerns: folk at the bottom are concerned with operational nitty-gritty issues; those in the middle worry about administrative issues; those at the top are focused on strategic issues. Thus information selected at the bottom for its operational relevance may be useless from a strategic perspective – especially after the middle managers have put their “administrative” spin on it. Furthermore, information that is relevant from a strategic perspective may never be noticed because it is irrelevant for operational purposes. The enhancing of information for strategic purposes is demonstrated by the use made by Britain’s Joint Intelligence Committee in the run up to the Iraq war. In that case, an unsupported piece of information – that Hussein could deploy WMD at 45 minutes notice – was bolstered to become a major plank in the British Government’s case for going to war. Chapter 6 of the Hutton report describes the process nicely.

Uncertainty absorption is probably the most difficult for the organization, especially a large differentiated organization, to solve. It involves people in different divisions and at different levels taking the time and energy to understand the mindsets of people in different positions in the organization. Presidents need to talk to operatives, marketers to researchers, and, in government, everyone needs to understand the political implications of what they do.

For each of these barriers to effective communication, Presidents and top managers need to be proactive. They need to actively explore frames of reference, they need to challenge the assumptions underlying the options that are proposed to them, and they need to get opinions from all parties affected by the decisions they make.

Image provided courtesy of: http://www.flickr.com/photos/deapeajay/1928521563/

Martin G Evans is Professor Emeritus of Organizational Behaviour at Rotman School of Management, University of Toronto