It’s too early to tell whether hiring an active CEO benefits the company doing the recruiting, because too few of these CEOs have completed their career at the receiving company. Because CEO tenure currently averages 7.9 years, the five consecutive years of data we have is too short a time to provide a complete picture of how well these retread chiefs have done. However, if the generally subpar results of prior CEOs hold true for active CEOs hired from other companies — and we think that’s likely — then “beggar thy neighbor” recruiting is a lose–lose: bad for the company and bad for the economy. The likeliest big winners will be the CEOs who move, not only because of their greater compensation but also because of the excitement of mastering new challenges.
Failed Apprentice Model
The third common succession strategy — shifting the chief executive to chairman of the board while promoting a second individual, from inside or outside, to the CEO position — provides another example of a seemingly prudent idea whose benefits are illusory.
This “apprentice model” covers 36 percent of the CEOs who departed in 2005. It is especially prevalent in Japan, where more than two-thirds of the CEOs departing in 2005 had served under a board chairman who was the former CEO. (See Exhibit 7.) In theory, the apprentice model sounds great: Not only does it satisfy the demand of governance activists to separate the role of chairman and CEO, but it keeps the skills and experience of the former CEO available, provides mentoring for the new CEO, and delays the extra responsibility of chairing the board for the new CEO until he or she is up to speed.
Unfortunately, logic fails in the marketplace, at least outside Japan. When we compared the three possible governance models — the combined CEO–chairman; distinct roles, with someone other than the previous CEO serving as chairman; and the chairmanship held by the former CEO — the best-performing companies were those in which the roles were split and the chairman was a true outsider, not the former CEO. This gap between insider and outsider chairman models is consistent in North America and Europe for most of the years we have studied.
The primary reason the apprentice model doesn’t work, we believe, is the inevitable division of responsibility and authority. No matter what the title and what the claims about the division of labor between CEO and chairman, the new chairman at these companies was the chief. For many years, everyone had looked to him. He had set direction for the company, controlled promotions and compensation, and embodied the company’s culture to both employees and external stakeholders. Now in his new position, he has tight links to existing executives, men and women who know how to reach him if they are unsettled by the successor’s strategy or actions. And everyone knows that if the former CEO is unhappy with either the direction of the company or its performance, he might fire the apprentice and take back the CEO title. Can an apprentice CEO realistically drive fundamental change that might well appear to be a repudiation of the past CEO’s actions, especially when disgruntled executives can complain to their former boss?
Moreover, the presumed benefits of retaining the former CEO as chairman are offset by negative effects. Having the former CEO around to offer guidance implies that the new CEO needs additional training and isn’t really qualified to do the job, undermining his or her authority. Letting the former CEO manage the board — a board whose members know the former CEO and were probably appointed by that CEO — hampers the new CEO’s ability to get full understanding and buy-in of his or her change agenda.