After X Years, the CEO Should Go

The thought behind this essay began in an unlikely, noncorporate place.

I was checking the proofs of the latest book by my wife, Meredith Hooper, published for the centenary this year of Capt. Robert Falcon Scott’s fateful last expedition. She describes how the death of Scott and his four companions returning from the South Pole (having been beaten to it by the Norwegian Roald Amundsen) overshadows and rewrites the whole story of the expedition: “Accounts of what happened shifted to take account of what was to happen. Writing, and thinking, accreted ‘hindsight bias’: understandable, unavoidable and insidious. No one wished to challenge the overriding narrative. To dislodge the tone and beat of the powerful and moving story that had swept into the imagination of a world audience.”

It suddenly occurred to me that chief executives sooner or later, after x years, inevitably fall into the trap of hindsight bias. Consciously or unconsciously, they become more concerned with justifying yesterday’s decision and creating the ”overriding narrative” of their legacy than with making the optimal decision for tomorrow. Yet, the best decision for the company’s health tomorrow may require acknowledging that yesterday’s decision was a bad one. At some point then, irrespective of performance, might not the CEO begin to be a liability?

Nonexecutive directors like me are, according to Britain’s Combined Code, probably past our sell-by date after six years: “Any term beyond six years for a nonexecutive director should be subject to particularly rigorous review and should take into account the need for progressive refreshing of the board.” Could this not be equally true of executive directors?

I was discussing this idea with a chairwoman recently, and she made the point that it is not just the individual CEO’s problem. She pointed out that over time the whole company “becomes lined up behind the CEO’s idiosyncrasies.” When I mentioned this to Joel Kurtzman, the editor in chief of Briefings, he immediately thought of the considerable amount of research devoted to Groupthink, the mode of thinking that happens when the desire for harmony in a decision-making group overrides a realistic appraisal of alternatives.
Hindsight bias overlaps with another form of bias. We process data in skewed ways. Information that supports existing views is welcomed, while that which does not is rejected. This phenomenon is known as confirmation bias.

If you accept that the CEO should go after x years because of the growing dangers of hindsight and confirmation bias and Groupthink, then what is the value of x? It is probably in the range of five to seven years — which is when nonexecutive board members in Britain are encouraged to move on. Among other reasons, it is thought that they become too close to management, losing objectivity and distance from the canvas.

I believe that the CEO should depart after a fixed term irrespective of performance. Boards and shareholders (not to mention CEOs) may find this a step too far. But there are a number of good reasons. Hindsight and confirmation bias are insidious. They do not happen suddenly; they creep up on the organization and may not become visible until it is too late. In managing the CEO’s expectations, the chairman and board may find it useful to state up front on appointment that the CEO should expect a term of around, say, five years. This obviates the need for difficult, surprising and embarrassing discussions with the CEO around Year 4.

There are two problems with my argument — one almost the inverse of the other. A predetermined tenure could lead a CEO to focus only on short-term goals, seeking to ensure that he or she is a hero after x years through profit-pumping. This could actually jeopardize the longer-term future of the company. It is often said that a great engine of the German economy has been the small and medium-size Mittelstand companies and that their success can be attributed to long-term thinking. Such companies are often family-owned, and the CEOs have clear motivation to leave the companies in great shape for sons and daughters in succeeding generations.

The second problem is known in the City of London as “kitchen sinking” and concerns the behavior of the new CEO. A new chief executive may be motivated to pour scorn on his predecessor and to initiate one-time accounting actions to clean up unsuccessful past actions. Such steps reduce profits and the share price, but improve the outlook for the CEO’s own share options and future heroism. A recurrence of this every five years would damage the accretion of shareholder value.

There is no obvious solution to this conundrum. Awareness of it would be a good start. All of us on boards should be actively and continuously aware of the power of hindsight bias, confirmation bias, legacy building and Groupthink in ourselves and in others.

Indeed, boards and chairmen should sharpen their skill at foresight bias. Niccolo Machiavelli got it right in 1532: All wise princes ”consider not only present but also future discords and diligently guard against them; for being foreseen, they can easily be remedied, but if one waits until they are at hand, the medicine is no longer in time as the malady has become incurable; hectic fevers ... at their beginning are easy to cure but difficult to recognize, but in course of time when they have not at first been recognized and treated, become easy to recognize and difficult to cure.”


Richard Hooper has devoted his career to the media and communications universe, having held senior executive posts in the BBC, government research and development, British Telecom, new media and satellite television. He was named a Commander of the British Empire by Queen Elizabeth II in 2006 for his service to the communications industry.