Warren Bennis: Share the Power
I am always amazed when leaders cling to strategies that don't work. The Walt Disney Co. is currently facing a revolt led by Roy Disney. His contention: that Disney CEO Michael Eisner has hurt the company by "consistent micro-management" and by his unwillingness to share power. Whatever the truth of Disney's charges, the perception that Eisner is a corporate tyrant running an underperforming company has helped to make both him and the company vulnerable, as evidenced by his loss of the chairmanship and the takeover bid by Comcast Corp.
Although command-and-control leadership has been under assault since the end of World War II, it continues to be widely practiced, usually to the organization's detriment. Almost everyone pays lip service to the notion that today's information-driven organizations need to be more collaborative and less hierarchical to be at their most effective. But collaborative organizations don't grow themselves. They are created, and the people at the very top are the ones who must insist on their creation. But the dirty little secret is that many leaders are uncomfortable with such structures. Many organizational leaders tend to cling to imperial roles, in large part, I suspect, because those roles feel marvelousthey provide the biggest jolt to the leader's ego. Sharing power and sharing informationthe prerequisites for collegial organizationsdo not enhance the top dog's sense of well-being. They only make all the others feel better.
There is a growing pile of evidence that more collaborative organizations are better adapted to solving the complex problems we face today. A group of social psychologists at MIT did an experiment decades ago that elegantly demonstrated the point. The researchers (I was a graduate student among them) built a round table, with partitions that separated the five subjects from each other's sight. Each subject was given six colored marbles. The subject's task was to determine which of his or her colored marbles was the same color as one of the marbles held by each of the others. To help make this determination, the subjects could pass messages on index cards through slots in the partitions.
The round table allowed us to simulate three organizational types. We called the first type, the one most like the standard bureaucratic pyramid, the Wheel. In that configuration, all the messages flowed to a single person, or leader. The most collegial configuration was called the Circle. In it, each subject could send messages to his or her immediate neighbors. A third type, called the Chain, allowed messages to flow in only one direction. As soon as a subject felt sure of the color of the target marble, the subject dropped his or her marble of that color down a tube in the table, thus allowing the experimenter to measure the accuracy and speed of the participants.
When the task was simplest, and all the marbles were a single-color type, the results were fairly predictable. The top-down Wheel was the fastest, most efficient organizational form. But when the task was made more complex, by using cat's-eyes and other variegated marbles, the collegial Circle was the fastest and most accurate, thanks to its opportunities for wider communication. And most important, perhaps, even when the top-down Wheel won, only its central leader reported feeling good about the experiment. The isolated nonleaders in the Wheel had little enthusiasm for the task. In contrast, on the more complex task, the Circle was not only superior, but all the participants also reported relatively high morale.
Time has confirmed these findings over and over again. Recently, I participated in a forum on 21st-century leadership in which the tension between tsarist leaders and the need for collaboration was a constant theme. Fellow participant Meredith Belbin, who has done extensive research on teams, offered an especially penetrating insight: He said that the alpha-male model of leadership may feel natural to us, as primates, but what today's world requires is more like the "sophisticated interdependent systems of social insects." Belbin argued that digital interconnectedness is driving leaders to change: "Information is coming in from the side instead of from the top down. . . . By losing its likely monopoly on leadership, the top can survive with credibility only by empowering the most suitable individuals and teams."
Surely changes in information flow are reshaping leadership. But I like to think that leaders are capable of changing themselves as well. One way they can encourage others to be collaborative is by doing it themselves. A handful already does. Michael Dell shares decision-making with the computer company's little-known president and COO Kevin Rollins. In June, Dell will also hand over his CEO title to Rollins. As Newsweek reported in its February 23, 2004 issue, the two executives removed the wall between their offices three years ago, and now their desks face each other, with only a glass partition in-between. Why is Dell able to share power when so many others cannot? He may be more collaborative by nature than most high-profile CEOs. Or he may simply know that an organization as complex as his needs coleadership. My guess is that the disparity in the men's ages makes the collaboration somewhat easier. Rollins is 51, and Dell, 39, may perceive Rollins less as a competitor than as an avuncular advisor.
A spirit of power-sharing can be implemented even in organizations that are assumed to be hopelessly hierarchical. I recently read Capt. D. Michael Abrashoff's book, It's Your Ship: Management Techniques from the Best Damn Ship in the Navy. One of the key ways the former commander of the destroyer USS Benfold was able to galvanize his once-dispirited 310-member crew was by ceaseless communication. He made sure that the crew knew what he was thinking, and why, and he urged his crew members to tell him what they really thought. Abrashoff made a fetish of walking around and talking to crew members on a daily basis. He communicated using every possible means, using the ship's public-address system so often that he became known as "Megaphone Mike." The more everyone knew, the better performance and the higher morale.
Indeed, Abrashoff's understanding of the importance of democratizing communication was a factor in both his success as a naval officer and in his contribution to his country. During the first Gulf War, operations were hampered by an enormous logjam of military messages that were often delayed, lost or never delivered. But in 1997 in the Persian Gulf, because the atmosphere on the Benfold eased barriers of rank, the ship's radioman, John Rafalko, was able to approach Abrashoff and suggest a solution. Rafalko's idea involved adapting underutilized rapid-data-transmission equipment, originally installed to help program Tomahawk missiles, to improve other secure communications. Rafalko convinced Abrashoff, who convinced his own superiors to allow Rafalko to teach his solution to other radio operators. Abrashoff reports the outcome: "Our first-class petty officer calmly updated all the gold-braided communications experts. . . [Using Rafalko's new system] the backlog problem disappeared virtually overnight. . . [and] ships suddenly communicated with one another around the clock, without a glitch."
To encourage leaders to share power and foster collaboration, their compensationeven their continued employmentcan be tied to it. But wholesale change will only come, I suspect, when leaders have a change of heart and discover what Abrashoff and Michael Dell have learned: Only when everyone in the organization believes that it's their ship can the organization truly thrive, and only then has the captain truly succeeded.
Warren Bennis is a Distinguished Professor of Business at USC and Chairman of Harvard University's Center for Public Leadership.