Decision making is an integrated sequence of activities that includes gathering, interpreting, and exchanging information; creating and identifying alternative courses of action; choosing among alternatives by integrating differing perspectives and opinions of team members; and implementing a choice and monitoring its consequences.2 For a schematic diagram of an idealized set of activities involved in a decision-making process, see Exhibit 7-1.

We begin by discussing how a variety of well-documented decision-making biases affect individual decision making and how these biases are ameliorated or exacerbated in groups. We identify five decision-making pitfalls that teams often encounter. For each, we provide preventive measures. We focus on groupthink, the tendency to conform to the consensus viewpoint in group decision making and then discuss escalation of commitment, the Abilene paradox, group polarization, and unethical decision making.

A variety of decision-making biases plague individual decision making. (For a comprehensive review, see Bazerman and Moore’s Judgment in Managerial Decision Making.)3 In this section, we briefly review four of the most well-documented individual decision-making biases and discuss their implications for teams.

A demonstrable task is a task that has an obvious, correct answer. Many management and executive education courses challenge businesspeople with simulations in which they find themselves stranded in inhospitable environments—arctic tundra, scorched desert, treacherous jungles—and together they must plan and enact strategies to ensure their survival.13 Some companies actually place people in such situations (see Exhibit 7-4). Popular television shows are based on “survival,” and some teams excel at adventure sports—but in the management laboratory and classroom, teammates are simply asked to rank (in order of importance) the usefulness of several objects (e.g., flashlight, canteen of water, and knives). The team’s rank order can be benchmarked against that of an expert and against the individual rankings made by each member of the team.

When this is done, an interesting phenomenon emerges: The performance of the team is inevitably better than the simple, arithmetic average of the group. Recall from Chapter 2 the team performance equation, wherein actual productivity (AP) of a team = potential productivity (PP) + synergy (S) – performance threats (T). In this case, the arithmetic average of the team represents the potential productivity of the group. If the actual productivity of the team exceeds this, it suggests that the group has experienced a synergistic process (i.e., working together has allowed the group to outperform how they could have performed by simply aggregating their own decisions). If the actual productivity of the team is worse, it suggests that the group process is flawed. However, the team leader is justified in asking whether the team’s performance exceeds that of the best member of the team. Perhaps team decision making is best served by putting the trust of the group into one knowledgeable and competent group member. One investigation studied individual versus group decision making in 222 project teams, ranging in size from three to eight members.14 In most instances, groups outperformed their most proficient group member 97 percent of the time.

Groups perform better than independent individuals on a wide range of demonstrable problems. For example, groups of three, four, or five people perform better than the best individual in letters-to-numbers problems (e.g., “A + D =?”)15, and groups outperform

individuals on estimation problems.16 Groups perform at the level of the second-best individual or group member on world knowledge problems, such as vocabulary, analogies, and ranking items for usefulness. Groups who use a structured approach for making decisions perform better than those without structure.17 People who have experience solving demonstrable problems in a group are able to transfer their performance to individual tasks,18 and people who anticipate group discussion are more accurate.19 Groups outperform individuals due to a process known as group-to-individual transfer, in which group members become more accurate during the group interaction.20

However, groups are much more overconfident than are individuals, regardless of their actual accuracy. For example, in one investigation, groups were asked to make stock price predictions. The actual accuracy of the group was 47 percent, but their confidence level was 65 percent.21 Three days before the disintegration of the space shuttle Columbia in 2003, NASA officials met to discuss ways to monitor and minimize the amount of falling debris during liftoff (one probable cause of Columbia’s breakdown), but concluded that repair work in flight—a possible solution—would be too costly, creating “more damage than what [they] were trying to repair.”22 Groups are also more likely to exacerbate some of the shortcomings displayed by individuals, namely, groups are more likely than individuals to (faultily) neglect case-specific information and ignore base-rate information.23

Groupthink occurs when team members place consensus above all other priorities—including using good judgment–when the consensus reflects poor judgment or improper or immoral actions.

Groupthink involves a deterioration of mental efficiency, reality testing, and moral judgment as a result of group pressures toward conformity of opinion. For examples of groupthink fiascos in the political and corporate world, see Exhibit 7-6. The desire to agree can become so dominant that it can override the realistic appraisal of alternative courses of action.30 The causes of groupthink may stem from group pressures to conform or a sincere desire to incorporate and reflect the views of all team members. Such pressure may also come from management if the directive is to reach a decision that all can agree to.31 Conformity pressures can lead decision makers to censor their misgivings, ignore outside information, feel too confident, and adopt an attitude of invulnerability.

Symptoms of groupthink cannot be easily assessed by outside observers. Three key symptoms of groupthink take root and blossom in groups that succumb to pressures of reaching unanimity:

• Overestimation of the group: Members of the group regard themselves as invulnerable and, at the same time, morally correct. This can lead decision makers to believe they are exempt from standards.

• Closed-mindedness: Members of the group engage in collective rationalization, often accompanied by stereotyping out-group members.

• Pressures toward uniformity: There is a strong intolerance in a groupthink situation for diversity of opinion. Dissenters are subject to enormous social pressure. This often leads group members to suppress their reservations.

Deficits arising from groupthink can lead to many shortcomings in the decision-making process. Consider, for example, the following lapses that often accompany groupthink:

• Incomplete survey of alternatives

• Incomplete survey of objectives

• Failure to reexamine alternatives

• Failure to examine preferred choices

• Selection bias

• Poor information search

• Failure to create contingency plans

Each of these behaviors thwarts rational decision making.

Coca-Cola’s decision to introduce New Coke in 1985 was eventually recognized as a mistake and reversed. Do such clear failures prompt teams to revisit their decision-making process and improve upon it? Not necessarily. In fact, under some conditions, teams will persist with a losing course of action, even in the face of clear evidence to the contrary. This is known as the escalation of commitment phenomenon.

Consider the following decision situations.

• As Facebook prepared for its highly anticipated IPO in 2012, inside sources said that both Facebook executives and Morgan Stanley bankers handling the offering ignored input about pricing and the value of the company. The initial offering was set at $38 per share. The stock dropped below that mark on the first day, before recovering to close just 23 cents above the initial price, a historically bad start for an IPO, which typically rises far above the initial price on the opening day. From there, Facebook began a drop that took it as low as $17.55 per share, putting many initial investors in a deep hole.52

• When Netflix suddenly announced that it would be separating its online streaming service and DVD mail service under the Qwikster name, they raised prices and outraged a loyal customer base. Complaints started immediately, but Netflix stood its ground. As subscribers dropped and Netflix’s stock price plunged, the company was forced to abandon the idea. The not well-thought out plan cost Netflix 800,000 subscribers.53

• General Motors (GM) put a lot of faith in the public appetite for electric cars, even speeding up the launch date for its Chevy Volt by 6 months. They were wrong. Despite company claims that the Volt was “virtually sold out” due to its popularity, the fact was that the car was hardly selling at all—only 125 Volts were sold in July 2011, worldwide. It wasn’t until the car went under investigation for fires involving the lithium-ion batteries that GM finally saw the writing on the wall.54

In all these situations, individuals and teams committed further resources to what eventually proved to be a failing course of action. Usually, the situation does not initially appear to be unattractive. The situation becomes an escalation dilemma when the persons involved in the decision would make a different decision if they had not been involved up until that point, or when other objective persons would not choose that course of action. In escalation situations, a decision is made to commit further resources

to “turn the situation around.” The bigger the investment and the more severe the possible loss, the more prone people are to try to turn things around.

The escalation of commitment process is illustrated in Exhibit 7-8. In the first stage of the escalation of commitment, a decision-making team is confronted with questionable or negative outcomes (e.g., a price drop, decreasing market share, poor performance evaluations, or a malfunction). This external event prompts a reexamination of the team’s current course of action, in which the utility of continuing is weighed against the utility of withdrawing or changing course. This decision determines the team’s commitment to its current course of action. If commitment is low, the team may withdraw from the project and assume its losses. If commitment is high, however, the team will continue commitment and continue to cycle through the decision stages. There are four key processes involved in the escalation of commitment cycle: project-related determinants, psychological determinants, social determinants, and structural determinants.55

The Abilene Paradox results from group members’ desire to avoid conflict and reach consensus at all costs.64 The Abilene Paradox is a form of pluralistic ignorance: Group members adopt a position because they feel other members desire it; team members don’t challenge one another because they want to avoid conflict or achieve consensus. The story in Exhibit 7-9 illustrates this dilemma.

It may seem strange to think that intelligent people who are in private agreement may somehow fail to realize the commonality of their beliefs and end up in Abilene. However, it is easy to see how this can happen if members fail to communicate their beliefs to each other.

Quandaries like the Abilene Paradox are easy to fall into. Strategies to avoid the situation include playing devil’s advocate, careful questioning, and a commitment on the part of all team members to fully air their opinions as well as respectfully listen to others. Note that none of these requires team members to abandon consensus seeking as a goal—if that is indeed their goal. However, it does require that consensus actually reflect the true beliefs of the team.

What factors lead to problems like the Abilene Paradox? In general, if individual team members are intimidated or feel that their efforts will not be worthwhile, then they are less likely to air or defend their viewpoints. This is called self-limiting behavior. According to a survey of 569 managers, there are six key causes of self-limiting behavior in teams:65

• The presence of an expert: If the group perceives that one of their members holds exceptional expertise or experience in the topic under discussion, individuals will self-limit their input.

  That July afternoon in Coleman, Texas (population 5,607), was particularly hot—104 degrees as measured by the Walgreen’s Rexall Ex-Lax temperature gauge. In addition, the wind was blowing fine-grained West Texas topsoil through the house. But the afternoon was still tolerable—even potentially enjoyable. There was a fan going on the back porch; there was cold lemonade; and finally, there was entertainment with dominoes. Perfect for the conditions.

  The game required little more physical exertion than an occasional mumbled comment,

  “Shuffle ‘em,” and an unhurried movement of the arm to place the spots in the appropriate perspective on the table. All in all, it had the markings of an agreeable Sunday afternoon in Coleman—that is, it was until my father-in-law suddenly said, “Let’s get in the car and go to Abilene and have dinner at the cafeteria.”

  I thought, “What? Go to Abilene? Fifty-three miles? In this dust storm and heat? And in an un-air-conditioned 1958 Buick?”

  But my wife chimed in with “Sounds like a great idea. I’d like to go. How about you, Jerry?” Since my own preferences were obviously out of step with the rest I replied, “Sounds good to me,” and added, “I just hope your mother wants to go.”

  “Of course I want to go,” said my mother-in-law. “I haven’t been to Abilene in a long time.”

  So into the car and off to Abilene we went. My predictions were fulfilled. The heat was brutal. We were coated with a fine layer of dust that was cemented with perspiration by the time we arrived. The food at the cafeteria provided first-rate testimonial material for antacid commercials.

  Some 4 hours and 106 miles later we returned to Coleman, hot and exhausted. We sat in front of the fan for a long time in silence. Then, both to be sociable and to break the silence, I said, “It was a great trip, wasn’t it?”

  No one spoke. Finally my mother-in-law said, with some irritation, “Well, to tell the truth, I really didn’t enjoy it much and would rather have stayed here. I just went along because the three of you were so enthusiastic about going. I wouldn’t have gone if you all hadn’t pressured me into it.”

  I couldn’t believe it. “What do you mean ‘you all’?” I said. “Don’t put me in the ‘you all’ group. I was delighted to be doing what we were doing. I didn’t want to go. I only went to satisfy the rest of you. You’re the culprits.”

  My wife looked shocked. “Don’t call me a culprit. You and Daddy and Mama were the ones who wanted to go. I just went along to be sociable and to keep you happy. I would have had to be crazy to want to go out in heat like that.”

  Her father entered the conversation abruptly. “Hell!” he said.

  He proceeded to expand on what was already absolutely clear. “Listen, I never wanted to go to Abilene. I just thought you might be bored. You visit so seldom I wanted to be sure you enjoyed it. I would have preferred to play another game of dominoes and eat the leftovers in the icebox.”

  After the outburst of recrimination we all sat back in silence. Here we were, four reasonably sensible people who, of our own volition, had just taken a 106-mile trip across a godforsaken desert in a furnacelike temperature through a cloudlike dust storm to eat unpalatable food at a hole-in-the-wall cafeteria in Abilene, when none of us had really wanted to go. In fact, to be more accurate, we’d done just the opposite of what we wanted to do. The whole situation simply didn’t make sense.

• A strong argument: If the group has spent a lot of time in circular discussions and idea fatigue has taken hold, a group member who presents a compelling and reasonable solution, even if it is not the optimal resolution, may find that their argument is met with agreement, because it allows the group to move on with business.

• Lack of self-confidence: When team members are unsure or uncertain about their contributions, they will self-limit.

• Trivial decision: When group members don’t see how the decision impacts themselves or something important, they will self-limit.

• Conformity: Roger Boisjoly, chief engineer on the tragic 1986 Challenger space shuttle accident, felt incredible pressure from the NASA management team to conform to their desire to launch.

• A faulty decision-making climate: When team members are easily frustrated and believe that others are dispassionate, involved, or apathetic, they are likely to self-limit. Such dysfunctional climates are unwittingly created at the outset of many team endeavors, when members complain and mock the process.

Consider the case in Exhibit 7-10. Most people independently evaluating the problem state that the new company would need to have nearly a two-thirds probability of success before advising Mr. A to leave his current job and accept a new position.70 What do you think happens when the same people discuss Mr. A’s situation and are instructed to reach consensus?

You might expect the outcome of the team to be the same as the average of the individuals considered separately. However, this is not what happens. The group advises Mr. A to take the new job, even if it only has slightly better than a 50–50 chance

of success! In other words, groups show a risky shift. After a group discussion, people who are already supportive of war become more supportive, people with an initial tendency toward racism become more racist, and people with a slight preference for one-job candidate will advocate for that person more strongly.71 In a field investigation that analyzed the decisions of federal district court judges, in the 1,500 cases where judges sat alone, they took an extreme course of action only 30 percent of the time, but when sitting in groups of three, extreme decision making doubled to 65 percent.72

Now consider a situation in which a company is deciding the highest odds of a drug contraindication that could be tolerated on the release of a new medicine. In this case, individual advisors are cautious, but when the same people are in a group, they collectively insist on even lower odds. Thus, they exhibit a cautious shift. Nearly 4 years after Merck & Co. took the painkiller Vioxx off the market because of links to heart attacks and strokes, the new regulatory climate altered the landscape of drug development. In 2008, the FDA approved just 19 new medicines, the fewest in the previous 24 years, and announced 75 new or revised “black-box” warnings about potential side effects—the agency’s strongest—twice the number in 2004.73

Why are teams both more risky and more cautious than are individuals, considering the identical situation? The reason for this apparent disparity has to do with some of the peculiarities of group dynamics. Teams are not inherently more risky or cautious than individuals; rather they are more extreme than individuals. Group polarization is the tendency for group discussion to intensify group opinion, producing more extreme judgments than might be obtained by pooling the individuals’ views separately (see Exhibit 7-11).

Group polarization is not simply a case of social compliance or a bandwagon effect. The same individuals display the polarization effect when queried privately after group discussion. This means that people really believe the group’s decision—they have conformed inwardly! The polarization effect does not happen in nominal groups. The polarization effect grows stronger with time, meaning that the same person who was in a group discussion 2 weeks earlier will be even more extreme in his or her judgment.

There are two psychological explanations for group polarization: the need to be right and the need to be liked.

Unethical decision making can be small scale or affect the lives and welfare of hundreds of thousands of people. Financier Bernie Madoff defrauded thousands of people by committing fraud, theft, and money laundering in a huge Ponzi scheme. At the time of his arrest, he claimed to manage $65 billion of investor money, but in reality there was just $1 billion.91 Thousands of people were affected and lost their entire life savings. Unethical decision making shares many of the same dynamics involved in the other concepts we have discussed in this chapter, such as groupthink. Groupthink can lead to a culture of unethical behavior within a company.92 Groups lie more than individuals when deception is guaranteed to result in financial profit.93 And, groups are more strategic than individuals in that they will adopt whatever course of action—deception or honesty—serves their financial interests.94 However, teams are concerned about ethics and value group morality more than competence or sociability.95 Below, we consider some of the situational triggers of unethical decision making. We resist the urge to explain all such occurrences of unethical decision making as a simple manifestation of evil personalities. Rather, we believe that certain conditions may act as enabling conditions for unethical behavior.