Leadership Research Summary:
- Using comprehensive data on 28 cohorts in Sweden, researchers analyze CEO health and its determinants and outcomes. The study find CEOs are in much better health than the population and on par with other high-skill professionals. These results apply in particular to mental health and to CEOs of larger companies.
- The study explored three mechanisms that can account for CEOs’ robust health. First, researchers find health predicts appointment to a CEO position. Second, the CEO position has no discernible impact on the health of its holder. Third, poor health is associated with greater CEO turnover.
- Here, both contemporaneous health and health at the time of appointment matter. Poor CEO health also predicts poor firm outcomes. Researchers find a statistically significant association between mental health and corporate performance for smaller-firm CEOs, for whom a one standard deviation deterioration in mental health translates into a performance reduction of 6% relative to the mean.
Leadership Research Findings:
- The study’s results have the following implications. First, the generally good CEO health and the lack of a discernible effect of CEO promotions and retirement on health suggest the anecdotes on the health crises of individual CEOs (some of which are listed in the first paragraph of the paper) likely are just anecdotes. Thus, there does not appear to be a widespread “CEO health crisis”. Second, the association between CEO health and turnover is strong, even when we account for publicly observable performance signals.
- Despite this relation, CEO health appears to matter for firm performance, at least in smaller firms. These companies may benefit from improvements in corporate wellness plans (Grobart, 2017), and the oft-neglected succession and contingency plans that assign emergency backups for the CEO (Bennedsen et al., 2020, Cheng, 2020).
- Relying on broad health indicatorsallows us to effectively summarize CEO health, its predictors, and associated outcomes. However, this approach comes with a price: researchers necessarily rely on associations rather than causal relations in most analyses.
- When a researcher’s setting does not lend itself to a causal interpretation, there is an attempt to mitigate problems associated with endogeneityand omitted variables with three different ways: (1) by employing a rich set of covariates, including CEO and firm fixed effects (for the CEO subsample) and early-life predictors (for the population), (2) by estimating our results with alternative specifications which employ both a rich and a scaled-down set of controls, and ascertaining the coefficients of interest do not change much between these specifications, and (3) by using lagged variables.
- Future analyses on CEO health may wish to establish causality by resorting to natural experiment designs (Sieweke & Santoni, 2020) that utilize, e.g., asymmetric, exogenous shocks to different industries. Such shocks could affect demands for CEOs and, indirectly, for their health. Another way to establish causality would be to study the effects of unexpected health shocks to CEOs themselves. Given the study’s evidence on the strong association between mental health and CEO outcomes, exogenous shocks to mental health (such as depression triggered by an unexpected loss of a loved one, Kristensen, Weisæth, & Heir, 2012) would be particularly interesting to study. Because well-defined unexpected health shocks are relatively rare, such analyses would likely need a large sample to reject the null hypothesis of no effect.
- Another interesting future research avenue relates to the corporate governance mechanisms associated with CEO health. Borgschulte et al. (2021)find that CEOs who serve under stricter corporate governance regimes face poorer long-term health outcomes. Currently, researchers do not know enough of the inner workings of firms to take a stand on the mechanisms linking health to CEO appointments, turnover, and performance.
- These mechanisms could plausibly be different, e.g., for firms representing different governance structures, for family and non-family firms, and for firms representing different levels of CEO ownership. To the extent boards (rather than owner-managers) have a decisive role in firm decision making, it is also not clear whether they view physical and mental conditions in the same way or whether mental conditions carry a stigma that makes CEOs suffering from them particularly vulnerable to board action.
- The fact that CEO outcomes are more strongly associated with mental health than physical health is consistent with a stigma explanation, but more research is needed to understand the stronger association of mental health with CEO and firm outcomes.