CEOs are made that way
Believe it or not, but CEOs are people, too. They have the same psychological strengths, weaknesses, and biases that everyone else has. And while we would love to think we live in a meritocracy where only the most skilled people become CEOs of companies, all it takes is one look at the business section of the newspaper to know that this is obviously not true.
So, how do CEOs become CEOs? That is the question that Ulrike Malmendier (who is one of my heroes because she produces some of the most fascinating research out there) and Marius Guenzel have investigated. In their paper, they look at three stages of the lifecycle of a CEO.
First, they look at the hiring stage. They show that overconfident people are more likely to be hired by boards as CEOs. To understand why, consider the case where a board wants to appoint an outsider to become CEO. The board is unable to observe the biases and character of the candidates directly because they haven’t worked with the person intensively and over a long time. What boards do naturally, in these cases is look at the track record of the candidates and how successful they were in their previous jobs. And most likely, the board will hire candidates that have managed to improve the value of their previous firms the most or that have demonstrated an ability to create a lot of added value in the projects they managed.
But what that does is to create a preference for overconfident managers. Overconfident managers systematically take on riskier projects. If the project is successful, the payoffs are larger and career advancement is guaranteed. If the projects fail, you drop out of the pool of candidates for future CEO job searches. Thus, there is a significant survivorship bias in the pool of CEO candidates that boards use to select CEOs. And as a result, newly appointed CEOs are more likely to be overconfident and excessive risk-takers.
Then, they looked at the actual job performance of a CEO. Once appointed, overconfident CEOs are obviously more likely to enter into riskier projects and pursue more aggressive financing decisions. The problem is that it is hard to establish cause and effect in investment and financing decisions so it is very difficult to directly blame a specific (positive or negative) outcome to the decisions of a CEO. This room for interpretation helps CEOs because self-attribution bias (which we all have) allows us to take credit for successes and blame failures on others or unforeseeable circumstances. Hence, CEOs don’t have a direct feedback mechanism that would allow them to learn from past mistakes and correct their overconfidence over time. What makes things worse is that learning opportunities typically are rare to begin with. Large investment projects (like M&A) or major financing decisions don’t happen every day. In fact they happen very rarely, making it impossible to engage in iterative learning on the job.
Finally, they show that because boards themselves are biased, the dismissal of CEOs doesn’t follow purely rational evaluations either. A board that has at least 40% women on board is 1.5 times more likely to fire an underperforming CEO than an all-male board. There is indeed something like the old boys’ club and it is often hard for boards to dismiss CEOs who they may be friends with or who are in many ways like themselves.
CEO, it’s a good job if you can get it…