Turbulence at the Top: Antecedents of Key Executive Dismissal

Article excerpt


Because key executive dismissal events are often played out in the media and serve as public signals of major organizational turning points, these events hold a special fascination for practitioners and scholars alike. The GM board coup, in particular, was a major symbolic event in the history of corporate governance. Not only has it precipitated other revolts by boards, but it has also pointed to the need to develop additional theory about what causes boards to wake up and take the hard actions. The stunning shake-up at GM has given a whole new meaning to the famous remark made by the late General Motors chairman, Charles E. Wilson, "what is good for General Motors is good for the country, and what's good for the country is good for General Motors" (Dobrzynski, 1992).

In this study we reexamine the antecedents of key executive dismissal. While this topic has been researched previously, we extend this research by limiting our study to those firms that specifically state that they are firing a key executive. Prior research has not used as specific a definition for executive dismissal. To be included in our sample the story in the press about the executive's departure had to have mentioned the word "fired" or "dismissed." By comparing this sample of firms that fired an executive to a matched sample that did not, we find that a firing generally requires both motive and opportunity. The motive appears to be poor financial performance, and the opportunity is the presence of a large number of outside directors with some potential pressure from debtholders. By using a more strict definition of firing than used previously and by reporting somewhat similar results as in past research, we confirm the findings of these earlier studies and show that they were not the result of misspecifying the definition of key executive dismissal.

Research Issues

Kesner and Sebora (1994, p. 366) in their review of over 30 years of succession research urge researchers to "continue their efforts to identify how the various succession variables fit together and where gaps in the literature exist." Rather than viewing succession in general, we focus on those events where a CEO was specifically dismissed. We pursue this line of research because top management ousters are a unique form of involuntary turnover whose antecedents likely differ from those of other involuntary separations, such as mandatory retirements (Friedman and Singh, 1989) and death (Worrell, Davidson, Chandy, and Garrison, 1986). Most prior research investigating top management change, however, has treated turnover as a general phenomenon, and only some researchers have attempted to model the variety of reasons of key executive change. Separating them should make it easier to develop viable theory and should improve academic understanding of the overall process (Worrell, Davidson, and Glascock, 1993). In the case of key executive dismissals, the board of directors presumably acts to correct a suboptimal match between top management behavior and the needs of the firm (Furtado and Karan, 1990). A dismissal exists when the departure is initiated by the firm, is not part of a formal policy such as in statutory retirement, and is against the executive's will. The principal objective of this study is to provide data concerning the major antecedents of key executive dismissal and to interface prior research in the finance and management literature.

For a key executive to be fired, several underlying conditions may be necessary. It is important to recognize that a firm's board of directors is elected by shareholders to represent the owners of the firm, among other things. The board is generally responsible for promoting or hiring the CEO and other key executives. The board is also responsible for removing a key executive who does not provide the results expected by shareholders. As long as the board members have the ability to remove the management, they have what Mizruchi (1983) referred to as "bottom line control. …