Matching, learning and social networking in CEO and director labor market
Abstract
How does a corporate board select their new CEO and directors? These questions are interesting to both practitioners and researchers. In my first essay, I develop a search-matching model which predicts that new CEOs who are better matches to the firms will stay longer, perform better, and require less initial compensation. Using comprehensive EXECUCOMP dataset merged with information on CEO succession plans, I find significant benefits to firms that adopt a CEO succession plan: insider successors in such firms have (1) 30% lower hazard rate of early turnover, (2) 10% higher productivity of effort, and (3) $1,000,000 less initial compensation. I also find that more independent boards (measured by the percentage of outside directors) and boards with larger social networks (measured by the average outside directorships of each director) pick CEOs with higher match qualities. All the findings suggest that the time boards spend on screening candidates and the size of boards’ social network play a role in CEO succession. My results also provide support and policy implications for a legal notice from SEC, dated October 27th 2009, which signaled increased concern about CEO succession planning among corporate boards and for the first time allows shareholders to request more disclosure from companies’ CEO succession plans. My second essay evaluates the impact of board interlock networks (i.e. boards are connected by interlocked directors–individuals who are officers and/or directors at two or more firms) on directorship market outcomes and firm values by estimating a dynamic structural model. The model features endogenous network formation and network learning and predicts that, under certain restrictions, the structure of board interlock networks can fully reveal previously unidentified objectives (e.g. whether the board is the firm value maximizers or social alliance to the CEO). The model’s parameters are estimated by applying Bayesian Markov-Chain-Monte-Carlo methods to longitudinal interlock network data of U.S. public traded firms over ten years (1996–2006). I find that social network connections are important in determining who gets which directorship. Specifically, the possibility of obtaining board seat increases by 30–150 times if the director has a tie to the hiring board, while other variables (e.g. director’s demographic characteristics, the performance of his current firm, his professional title, his compensation, how often the director attends a board meeting, etc.) have negligible effects. I also find that boards discount the value of directors with a large number of outside directorships, indicating a desire for effective monitors. Finally, I find that interlock connections decrease firms’ values pre-Sarbances-Oxley Act (SOX) and increase post-SOX. Overall, this model provides the first study in the dynamic relationship between board interlock network, director appointment and firm value.
Degree
Ph.D.
Advisors
McConnell, Purdue University.
Subject Area
Management|Economics
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.