The benefits and drawbacks of generalist CEOs – those with the talent and skill to manage companies in various industries – have been fiercely debated in contemporary research. Some research touts generalist CEOs’ varied professional experiences and ability to launch a wide range of strategic initiatives. Other research warns of their tendency to switch jobs easily, which may mean that their motivations do not align with those of shareholders and prompt them to give short-term investments priority over above longer-term, but more beneficial, projects.
Our research contributes to the debate by exploring how independent directors view generalist and specialist CEOs. Boards of directors are often considered the ultimate governance mechanism for resolving agency conflicts, and, as outsiders, independent directors are more likely to be impartial. As a consequence, the degree of board independence is often used as a measure of board quality. Since effective board governance leads to shareholder-friendly corporate decisions and activities, a great deal of research demonstrates that independent directors are valuable. Therefore, to the extent that generalist CEOs enhance shareholders’ wealth, independent directors should view them favorably.
It is, however, challenging to test this argument empirically because, in economics and finance, it is not possible to run a randomized experiment where certain firms are assigned more independent directors than others are. Fortunately, there is a quasi-natural experiment where only certain firms are forced involuntarily to change their board independence by a regulation. This is analogous to a randomized experiment. Interestingly, three economists have just been given a Nobel Prize for their work related to this empirical method (David Card, Guido Imbens, and Joshua Angrist).
In 2002, the Sarbanes-Oxley Act, along with new stock exchange rules, required that publicly traded firms have a majority of independent directors on the board. This requirement served as an unexpected exogenous shock to companies who had to appoint more independent directors to comply. Using a difference-in-difference estimation, we compare, before and after the passage of the law, the change in the general ability of the CEOs of firms forced to appoint more independent directors to the change in the general ability of the CEOs of firms not affected by the law. Firms required to change their board composition were the treatment group, and firms unaffected by the law were the control group.
Our difference-in-difference estimates reveal that firms compelled to increase board independence show a smaller improvement in CEO ability than do other firms, suggesting that independent directors do not view generalist CEOs favorably. Our findings are consistent with the notion that generalist CEOs may have incentives that conflict with those of shareholders, resulting in an incentive misalignment. For instance, generalist CEOs are less likely to be risk-averse (Custodio et al., 2013; Mishra, 2014). Supporting this argument, Mishra (2014) reports that having a generalist CEO leads to a significantly higher cost of equity capital. Moreover, because generalist CEOs tend to change jobs more often, they may not take the long-term perspective required to enhance shareholder value. Furthermore, our results are consistent with those in Ma, Ruan, Wang, and Zhang (2021), who find that companies with generalist CEOs have significantly lower credit ratings, implying that credit rating agencies view generalist CEOs as a credit risk factor. Their results are consistent with ours. Both independent directors and credit rating agencies view generalist CEOs unfavorably.
We also run several checks to ensure that our findings are robust. For instance, we execute propensity score matching, where we carefully match each treatment firm to another firm outside the treatment group that is most similar based on several firm characteristics. So, our treatment and control firms are nearly identical except for board independence. This technique increases the probability that the findings are driven by board independence, not by any other firm characteristics. Moreover, an instrumental variable analysis, which is more likely to reveal a causal effect, corroborates our findings. Finally, we apply Oster’s (2019) method for testing coefficient stability and find that our conclusion is robust.
Our results contribute to the debate about generalist CEOs. Most prior studies concentrate on assessing the effects of generalist CEOs on corporate outcomes and policies. Our study, however, adopts a unique approach, focusing on how independent directors view generalist CEOs. Furthermore, our study contributes to the literature that exploits the Sarbanes-Oxley Act to ascertain the effects of board independence on various corporate outcomes. Our study is the first to apply this approach to explore the impact of board independence on general managerial skills.