Should more CEOs be fired?

    New research suggests that more CEOs would be shown the door if not for their personal ties with their boards, suggesting that it's time to rethink hiring and firing practices and the CEO's role at public companies.

    Boards of directors are facing increasing pressure to fire CEOs rather than allow them to linger. But what about the pressure boards feel to keep their CEOs in place?

    A recently published study by Wharton professor Luke Taylor shows that boards of directors consider things beyond immediate costs (such as severance) when deciding whether to fire the CEO. Whether boards really consider such immediate costs very seriously at all is debatable.

    A recent case in point? Google's (GOOG) board gave CEO Eric Schmidt a $100 million option and stock grant as a going away present.

    The real business motivation of this move is elusive. These kinds of awards hurt shareholders. Schmidt already owned over $5 billion in Google stock. Even if the board didn't care about the impact on shareholders or other stakeholders, would stock and options be the best choice as a parting thank you?

    Schmidt is even trying to get rid of some of his stock by setting up a plan to sell approximately $335 million of shares he already owns.

    Of course, Google isn't alone in bonanza giveaways of shareholder assets when CEOs leave. By historical standards, Google's giveaway competes with those where the CEO was clearly forced out.

    Since Schmidt moved to the executive chair role at Google, he would not be considered as "forced out" of the CEO role, according to the criteria used in Taylor's and similar studies. Those companies with that allow the CEO time to find another job (in or outside the company) or to retire with six months notice, for example, wouldn't make the "forced out" criteria, either. This "probably underestimates the rate of forced CEO turnover," Taylor argues in his study.

    Using this definition, on average, 2% of the CEOs at the largest 500 companies were forced out each year between 1970 and 2006. But Taylor argues that this number would be much higher if directors cared more about shareholder value or were not so loathe toward forcing CEOs out, largely for personal reasons. These personal reasons may include their own ties to the CEO, or considerations that firing the CEO may put their jobs as directors at risk, or hurt their chances of being nominated to other boards.

    Consider the recent case at HP (HPQ). Former CEO Mark Hurd was forced out by the board for ethics violations. For the board members, however, removing Hurd was a courageous act. That move, in fact, put the directors' jobs at risk. The new CEO, Leo Apotheker, with the aid of new director and chair Ray Lane, asked for directors to volunteer their resignation so that they could rebuild the board with directors that they both knew.

    The model developed by Taylor predicts that instead of 2% per year, the percentage of CEOs forced out per year would be closer to 13% if boards were not influenced by these personal factors. (A 1999 study published in the Quarterly Journal of Economics showed that 15.5% of all mutual fund managers are forced out each year.)

    If more CEOs were dismissed, shareholder value would increase, according to Taylor's model. But Taylor also notes that it is difficult to say "how much is optimal."

    As one might expect, boards with fewer outsiders and boards at smaller firms are more heavily influenced by their personal relationships with the CEO when making firing decisions, according to Taylor. Personal factors also seem to have had a bigger influence further in the past (1970 – 1989) compared with more recently (1990 – 2006).

    Even with a very limited definition of what constitutes a forced CEO dismissal and a large gap between the actual forced dismissal rate and what theoretically might happen if a board puts aside its personal barriers to firing a CEO, quite a few CEOs are forced out of their companies. While 2% are forced out every year, 17% of all CEO tenures ended in forced dismissal from 1970 to 2006.

    Between 2000 and 2006, forced dismissal rates were even higher, averaging nearly 4% annually, with over one in four CEO tenures ending in forced dismissal. However, 4% per year and one in four stand in contrast to the fate suffered by shareholders: an actual decline in the S&P 500 during this period.

    The correlation between CEO dismissal and the performance of a company's stock suggest that there are serious issues with boards' firing practices, and more importantly, their hiring practices.

    The stats make it clear that it's time for directors to review their hiring and succession practices and, more broadly, rethink the CEO position and its role. And they may need to consider a different approach, perhaps even a team approach, to the job.

    Do CEOs matter too much? CEOs command large salaries, particularly in the largest firms, but is this warranted? If boards don't believe it matters whether a CEO stays or goes, should other stakeholders care? What is the risk that any given board will choose a worse option the next time around?

    According to Taylor, the larger the company, the less a CEO's performance seems to differ from his or her CEO peers. One explanation is that CEOs in larger firms have less of a direct impact on the company's performance -- perhaps because they delegate more than CEOs at smaller firms. So does replacing one CEO with another, in the largest firms, really produce a significant difference in results?

    This is something Bob Benmosche, CEO of AIG (AIG), hinted at in a November 15 interview in the Wall Street Journal: "I disagree that it would be a setback for AIG if I have to stop working. This company is here not just because of me."