Mechanisms of Board Turnover: Evidence from Backdating

In our paper, Mechanisms of Board Turnover: Evidence from Backdating, which was recently made publicly available on SSRN, we examine a set of events that involve observable corporate misdeeds: stock option backdating. These misdeeds were generally revealed within a narrow window of time, required the complicity of the board, and in many cases directors benefited directly through backdated grants. Examining board turnover associated with stock option backdating thus enables us to gain more insight about the mechanisms by which directors depart their boards. Although information that would allow us to identify each of these five steps is not publicly available, events that are typically available include the following: (1) whether a director resigns, (2) whether a director appears on the proxy as nominated for reelection, and (3) whether a director is reelected. Additionally, there are press releases that sometimes accompany these decisions, but these announcements must be interpreted with care. For example, when a director does not appear on the proxy, the board and/or nominating committee might have chosen not to renominate the individual for reelection or the director might have declined to stand for reelection (an event that is frequently disclosed, especially if driven by a director retirement policy). However, a director who will not be renominated often is permitted to announce that he or she has chosen to resign or not to seek reelection.

To examine the details of directors’ departures, we use a hand-collected sample. Consistent with directors being disciplined for fraud, litigation, and/or mismanagement (Gilson, 1990; Gerety and Lehn, 1997; Ferris, Jandik, Lawless, and Makhija, 2007), we find increased director turnover following the disclosure of backdating. Importantly, we find that most director turnover occurs prior to the election date. We find that inside directors’ departures generally are due to resignations, whereas independent directors’ departures primarily occur from failure to secure renomination. We also document a reduction in the number of additional board seats held by directors who are implicated in backdating. Prior research documents director turnover and the loss of additional outside directorships in response to fraud and mismanagement. These papers generally focus on Fama’s (1980) notion of ex-post settling up for directors. However, these studies fail to examine the relation between changes in board composition and the board-election process.

In response to concerns that the limited number of contested elections is evidence of directors being ineffectively monitored or disciplined by shareholders, the SEC adopted rule 14a-11 regarding “proxy-access,” (on July 22, 2011 the U.S. Court of Appeals rejected this rule, following a challenge filed by the U.S. Chamber of Commerce and Business Roundtable). Criticisms of the SEC’s proxy-access rule largely focus on its potential ineffectiveness (Kahan and Rock, 2010), the effects of the rule on the type of directors that could be elected, the economic advantages to maintaining the centralization of decision-making in the board of directors, or the view that reforms to proxy-access rules are driven by political considerations and are unrelated to improving shareholder value (Grundfest, 2010). None of these criticisms considers the degree to which effective disciplining and monitoring of directors already occurs.

Our evidence implies that directors are rewarded or punished for their actions. Our findings suggest that board departures associated with misdeeds occur largely prior to elections—in part due to the threat of losing reelection. These results suggest that recent proposals to change proxy-access rules address an issue that is less significant than might otherwise appear. The loss in outside board seats for directors who are associated with mismanagement potentially will be stronger in the future. The Securities and Exchange Commission (2009) has adopted new regulations requiring additional disclosure of certain legal proceedings against a director nominee, extending the period from the past five years to the past ten years. Moreover, RiskMetrics (2009) has changed its policy, now recommending that their clients vote “against” or “withhold” for directors (individually, on a committee, or the entire board) due to, “egregious actions related to the director(s)’ service on other boards that raise substantial doubt about his or her ability to effectively oversee management and serve the best interest of shareholders at any company.”

The full paper is available for download here.