In the paper Reputation Penalties for Poor Monitoring of Executive Pay: Evidence from Option Backdating, forthcoming at the Journal of Financial Economics, my co-authors (Yonca Ertimur of Duke University and David Maber of the University of Southern California) and I examine whether directors are held accountable for poor monitoring of executive compensation.
Theoretical and empirical work suggests that outside directors incur reputation penalties in the director labor market for poor monitoring. However, it is unclear whether these penalties extend to poor monitoring of executive pay. A widely held view—articulated by Prof. Bebchuk and Prof. Fried in their book Pay without Performance—is that there is little or no accountability for excessive or abusive pay practices. Yet no study has empirically examined this question. Part of the reason is the difficulty of defining and identifying “poor monitoring” with respect to executive pay. In most cases, pay levels and structures can be justified on economic grounds (e.g. retention, incentives, attraction of talent) and with reference to the behavior of peer firms. Unless these practices are perceived as clearly “outrageous,” it is unlikely that directors will be concerned about reputation costs. Opacity in pay disclosures makes it even more difficult to assess the quality of pay practices.
Our study identifies the option backdating scandal of 2006-2007 as a powerful setting to empirically examine whether outside directors are held accountable for poor monitoring of executive compensation. There is little doubt that revelations of the backdating practice sparked outrage and that at least some of the BD cases were blatant incidences of “stealth” compensation. BD is also an interesting setting in that there is a time lag between the time it took place and the time it was discovered—thus, board and committee composition may have changed, allowing for tests of “memory” in the director labor market.
Using a sample of approximately 180 backdating firms, we analyze the reputation penalties incurred by outside directors subsequent to the discovery of backdating. As proxies for reputation penalties, we use votes withheld from directors at elections, director turnover, and the number of seats gained/lost on other boards.
We find that the percentage of votes withheld from directors at backdating firms is twice the percentage withheld from their counterparts at non-backdating firms. This voting penalty is more pronounced for directors sitting on the compensation committee, particularly those sitting on the compensation committee during the backdating period, suggesting that shareholders view backdating primarily as a monitoring failure of the compensation committee, rather than the audit committee, even though backdating also resulted in overstated profits. The voting penalty is significantly more pronounced in the most severe cases of backdating, with directors at such firms receiving 27% more votes withheld than their peers at non-backdating firms (that is, about six times more).
Similar to prior studies, we find a strong correlation between shareholder votes and the recommendation of Institutional Shareholder Services (ISS), the influential proxy advisor. However, our tests suggest that shareholders do not mechanically follow proxy voting recommendations. The percentage of votes withheld from directors who received a backdating-related withhold recommendation because they sat on the compensation committee during the backdating period is significantly larger than for directors who received a backdating-related withhold recommendation for failure to take remedial actions after the discovery of backdating, suggesting that shareholders cast their votes depending on the specific reasons for, and context of, proxy advisors’ recommendations.
With respect to director turnover, we find that the probability that compensation committee members at backdating firms lose their seats is 1.37 times higher than for compensation committee members at non-backdating firms. The turnover penalty is stronger when backdating is more severe and in poorly performing firms.
Finally, we examine whether directors of backdating firms experience reputation penalties at other, non-backdating firms. Interestingly, we do not find that directors of backdating firms receive more votes withheld when up for election at another firm or that they are more (less) likely to lose (gain) another directorship. Consistent with these results, we find no instance of ISS recommending withholding votes from a director at a non-backdating firm because of involvement in backdating at another firm (only recently ISS has begun to take account directors’ performance at one firm in making recommendations when they are up for election at another firm).
Overall, our study offers two key conclusions. First, involvement in backdating results in significant penalties for compensation committee members at backdating firms but not at other firms, casting doubts on whether such penalties are likely to have any ex ante effect on directors’ incentives to monitor CEO pay. Second, voting shareholders carefully distinguish different types of responsibility and circumstances surrounding backdating and do not blindly follow proxy advisors’ recommendations, suggesting that concerns with reforms empowering shareholders’ votes and with the influence of proxy advisors may be overstated.
The full paper is available here.