When the boss’ son takes the helm at work, it’s bad news for employees who were vying for the big promotion. But is it bad for the company? Economic intuition offers two plausible and opposing answers. On the one hand, a scion CEO probably has a large ownership share and therefore huge incentives to run the company well. He also knows a lot about the firm and faces the scorn of his extended family, as well as more distant critics, if he fails. On the other hand, his selection may reflect a CEO search that is less than thorough: The best candidate among the boss’s relatives may not be as good as the best available. A new study resolves this tension against the heirs-apparent, showing that if the new CEO is related to his or her predecessor, the firm’s performance will suffer.
The question of how installing the boss’s kid as CEO affects firms’ performance is actually quite hard to answer. An initial step is to look at how a company’s performance changes after a CEO succession, comparing firms with scion CEOs with those in which executive power passes outside the family. The authors of the new study do this, and they show that in a sample of 5,000 Danish firms from 1994 to 2002, firm performance (measured by the ratio of operating income to assets) improved only after a nonrelative took over. The ratio of operating income to assets averaged 3.3 percent in their sample—for a firm with a million dollars in assets, that puts annual operating income at $33,000. Company performance improved 1.3 percentage points ($13,000 in annual income per million in assets) after the succession of outside-the-family CEOs. And it declined 0.1 percentage points ($1,000 in annual income per million in assets) when scion CEOs were chosen.
On its face, this looks like a straightforward measure of the effect of choosing unrelated vs. related successor CEOs. If this were the result of a controlled experiment, in which we could randomly select some firms to pass the torch to the boss’s relative and others to an unrelated candidate, then we would conclude that choosing Junior reduces operating income by $14,000 annually per million dollars in assets.
But the decision to hand over the reins within the family is not random. Firms may choose scion CEOs when calm seas are expected and a highly qualified outsider when storms are on the horizon. As with so many empirical questions about business, we are at a loss in uncovering causes as opposed to associations. Or rather, we seemed to be. The authors of the new study, Morten Bennedsen and Kasper M. Nielsen of the Copenhagen Business School, Francisco Pérez-González of Columbia University, and Daniel Wolfenzon of New York University, came up with a way to measure the effects of scion succession that solves this riddle. They hypothesize that firms whose male CEOs have male firstborns are more likely to hand over the reins within the family—in other words, they want to pass down control but are sexist.
The researchers get to test their thesis by taking advantage of Denmark’s relaxed attitude toward data disclosure. Italy is relaxed about work. France is relaxed about head-butting by football heroes. And Denmark is relaxed about giving researchers access to detailed personal data. The Danish government maintains a Civil Registration System with names and identification numbers of family members. Danish financial-performance data include names of executive officers, along with their personal identification numbers. By merging the two databases, one can determine which CEOs have family ties to their predecessors (both ties to sons that are easy to see because of a shared last name, and less-obvious ties to nephews, sons-in-law, etc.). Using these data sources together, the authors examine the effects on performance of the 5,000 Danish CEO successions.
The authors’ hunch about CEO sexism turns out to be right. When the outgoing CEO’s firstborn was male, succession passed within the family 40 percent of the time; when the firstborn was female, the in-family succession rate was only 30 percent. The sexism of the CEO dads produces conditions tantamount to the experiment we’d want to design. A random event—a firstborn boy—raises the probability of within-family succession from 30 percent to 40 percent. And here’s the juicy result: Firms in which the CEO dad had a male firstborn, a factor that by itself should have no effect on the firm’s subsequent operating performance, experience a $10,000 larger deterioration in income per million dollars in assets after a succession. That is, an “experiment” that raises the probability of within-family succession by 10 percentage points (from 30 percent to 40 percent) reduces operating income by $10,000 per million dollars in assets. Intuitively, this means that if the choice between a relative or nonrelative CEO were made by the flip of a coin, the choice of a scion CEO would reduce performance by about 10 times as much, or about $100,000 in operating income per million in assets. The authors conclude that professional management at the top—drawn from a large outside talent pool—is “extremely valuable.”
Like all studies, this one has limits—chiefly that it is limited to Danish firms. But we are left with the conclusion that showing favor to family members in the executive suite isn’t bad just for other CEO candidates. It’s bad for firms’ performance as well. To paraphrase the Bard, something nepotistic is rotten in the state of Denmark.