Speed of CEO Dismissal: An Attribution-Based Model of When Boards of Directors Fire CEOs in Response to Performance Downturn

Kim, Young Un

Download PDF

Request Version for Screen Reader

Last Modified

March 21, 2019

Creator

Kim, Young Un
- Affiliation: Kenan-Flagler Business School

Abstract

This research examines the dynamics of the CEO retention and dismissal decision making process adopting an attribution theory perspective. Replacement of management is generally known as an essential element for firms experiencing performance downturn and in need of organizational turnaround. However, firms vary in their speed of CEO replacement as a response to performance downturn, and the relationship of speed to the efficacy of response has not been examined. This dissertation investigates three broad research questions. First, how does causal reasoning based on performance feedback explain the variance in CEO dismissal timing? Second, do boards misattribute the cause of performance downturn inappropriately to the CEO? If so, what are the post-succession performance consequences? Last, are firms dismissed faster due to increased level of legal scrutiny in the post Sarbanes-Oxley era? While current debates about CEO dismissal have generally been dominated by economic and political perspectives on CEO/board relations, I argue in this paper that CEO dismissal may be driven by cognitive, behavioral, and symbolic reasons as well. I specifically examine how attributions of causality of different types of performance downturn affect the corporate boards' interpretations of CEO skill and their speed of response action. I predict that if board of directors view downturn as being internally caused and permanent, then the CEO will be dismissed faster. If boards view the downturn as externally caused and temporary, then the boards will be less likely to dismiss the CEO or be late in their dismissal actions. Based on a sample of 376 CEO dismissal observations in 348 public US companies during the period 1992-2009, I find that firms with a moderate speed in CEO dismissal outperform those that have faster or slower speeds of dismissal, showing that the response time after a downgrade can be an important variable affecting firm performance. The analyses also provided clear support for the relationship between different types of performance downturn and speed of CEO dismissal. Last, results also show that the time it takes to fire a CEO has increased over time (slower CEO dismissals), especially after the Sarbanes-Oxley legislation.

Date of publication