Pulpit Bullies: Why Dominating Leaders Kill Teams

When Harvard Business School Associate Professor Francesca Gino invites high-powered business leaders to address her class, she often observes an interesting phenomenon. The guest speakers announce that they are just as interested in learning from the students as teaching them, and encourage them to ask questions and make comments. In reality, however, the speakers often do the opposite—dominating the time and not allowing for much discussion at all.

"As professors we do this too," admits Gino. "It's very difficult when think you have the right answer not to put it out there." At the same time, she has observed, by hogging the discussion, these leaders not only limited their own learning but also made the class less productive as a whole.

Gino wondered if the same dynamic could be occurring in business, with dominating leaders stifling creative ideas that might otherwise emerge from group discussions and making the teams less productive.

The observation would run counter to the way we usually think about group dynamics—surely, a strong leader naturally improves the functioning of the team. Together with Leigh Plunkett Tost of the University of Michigan and Richard P. Larrick of Duke University, Gino explores this question in When Power Makes Others Speechless: The Negative Impact of Leader Power on Team Performance, forthcoming in the Academy of Management Journal.

In a series of studies, they found that when leaders were focused on their own sense of power, they can hurt the performance of their teams—but with an important catch. The effects only occur when leaders are actuallyin a position of power.

Gino and her colleagues differentiate between a "subjective sense of power," when someone believes they have control over others, and actual power, when someone has formal authority over how resources are allocated or how decisions are made. The two often go hand in hand, but not always. Sometimes in a group situation without a formal leader, for example, a leadership role can be assumed by a person who believes he or she has superior knowledge or skills.

UP THE MOUNTAIN

Gino, Larrick, and Tost tested this dynamic in a simulation involving people planning an imaginary climb up Mount Everest. The team consisted of different specialists—including a professional climber, a doctor, and a photographer—each of whom scored points according to how many of their individual goals were met.

For each group, the researchers designated a formal leader. In some cases they created feelings of power by asking the leaders to write about a time when they held control over others; in other cases, they didn't.

The results were striking. The "high-power" leaders who had done the writing exercise dominated the discussion, talking for 33 percent of the time, while the "neutral-power" leaders talked almost half as much, 19 percent. As a result, the first group of leaders missed important clues, such as information from the doctor about the oxygen running low or opportunities by the photographer to earn more points if they stayed an extra day at a certain base camp.

In those cases, the team as a whole suffered as well, achieving an average of 59 percent of the goals in the first group, compared to 76 percent in the second group.

"Even subtle ways of making people feel powerful have powerful effects on behavior," concludes Gino.

In a separate study that tested a group's ability to solve a murder mystery, however, Gino and her colleagues found that individuals in some situations aren't always so easily cowed. For this study, in which individual team members held different clues essential to solving the mystery, the researchers used two variables—in some cases, appointing a formal leader and in some cases not; and in some cases, priming individuals to feel powerful and in some cases not.

They found that in cases when someone felt powerful but was not recognized as being in a position of authority, team members were able to override that person's domination of the conversation and add their own input. Of the four groups, the two without a formal leader had the same performance, solving the mystery about 60 percent of the time. But the groups with a formal leader who was also primed to feel powerful did the worst, getting the right answer only about 25 percent of the time.

Surprisingly, however, the best groups were those who had formal leaders not influenced ahead of time to feel powerful—they solved the mystery nearly 80 percent of the time.

"My sense is that the leader is sort of stepping back," says Gino. "It's more of what you like to see, where the leader is orchestrating the conversation, but everyone is talking."

In other words, strong leaders can improve team performance, but only when they go into a situation with a sense of humility about their own relative power.

REINFORCING THE MESSAGE

The researchers expanded on that point in the last study, in which participants were asked to play the role of a management team tasked with advising the CEO on which CFO candidate to hire. In this case, the researchers appointed a formal leader for each group and primed some, but not all, to feel powerful. In half of the cases, however, they also performed a simple intervention, reminding the group leaders that each participant had unique insights to contribute.

In cases when this intervention occurred, the groups with high-power leaders did the best, coming up with the right answer an average of 60 percent of the time, slightly better than neutral-power leaders without intervention (56 percent) and neutral-power leaders with intervention (50 percent).

All three groups, however, blew away the high-power leaders who lacked intervention reminding them to listen to others—getting the answer 0 percent of the time. That suggests a powerful opportunity to improve performance just by making leaders aware of the dangers of hogging airtime in a discussion.

"I want to believe that oftentimes we behave the way we do because we are not aware of the effects of our actions," says Gino. "Bringing this type of awareness to leaders walking into group decision-making situations could set up a different process whereby they benefit from what others have to offer."

These findings don't let non-leaders in groups off the hook, however. Being aware of the negative effects generated by an overpowering leader can make non-leaders feel more empowered to assert their own point of view—whether or not the person dominating the conversation is a formal leader.

LIFE AND DEATH

The Gino team has started a new study, observing group decision-making situations in the field. Focusing on the health-care industry, they are experimenting with interventions to help improve outcomes in situations in which doctors—never known to be shy about exercising power—collaborate with medical colleagues in situations with real-life effects on patients.

If anything, the results should be more dramatic than the earlier experimental studies, where prompts were used to create feelings of power in group members. In these real-life situations under study, the doctors come fully equipped with a sense of power ingrained in them by having wielded authority over others for years.

Getting leaders to listen to others and to facilitate a productive group discussion in those circumstances would be powerful indeed.

Harvard Working Knowledge