Join our community of smart investors

Costs of overconfidence

Businesses can expand too quickly, even if bosses are not overconfident
March 8, 2018

Several restaurant chains are closing branches. This is the result of a common boardroom mistake – to expand too fast, invest too much or, in other cases, to embark upon too many takeovers. Why do they do this?

It’s widely thought that one reason is hubris. As Ulrike Malmendier and Geoffrey Tate have written: “a substantial share of top corporate executives exhibit symptoms of overconfidence in their decisions”. But there might be something else, pointed out in a new paper by Venkataraman Bhaskar and Caroline Thomas at the University of Texas in Austin.

Leaders, they say, stick with expansion plans in part for fear of looking weak or stupid if they reverse course. This stigma will be even greater than usual if the leader is thought to be overconfident. His underlings, investors and rivals will think: “his original decision must have been truly stupid if even that bumptious oaf now realises it was”. The fear of such a reaction might cause a boss to stick to his plans even if he is in fact not overconfident at all. This is especially the case because the evidence to warrant abandoning those plans will usually be noisy and inconclusive.

In this sense the problem is not necessarily that bosses are overconfident but that they are thought to be. Bhaskar and Thomas say: “A culture where leaders are expected to be overconfident can have adverse consequences even when the leader in question is not, in fact, overconfident.”

In fact, things might be worse. This assumes that the CEO gets information that might cause him to rethink his plans. But he might never get it. If junior managers believe their seniors are overconfident and so don’t want to hear bad news, they won’t tell it them. Even if those bosses are not in fact overconfident, they will then lack the feedback they need: some observers believe this has been one problem at General Electric. As Kenneth Boulding wrote in 1965, there’s a danger that leaders in large authoritarian organisations “will be operating in purely imaginary worlds”.

What can be done to stop this? One thing is for organisations to have mechanisms whereby bad news can be transmitted upwards, perhaps anonymously. 

Another possibility, say Bhaskar and Thomas, is to recruit leaders who don’t have a reputation for overconfidence and so don’t need to fear a stigma from changing policy when necessary. This, they say, suggests a case for a bias towards female bosses as women on average tend to be seen to be less overconfident than men. There is some evidence this might work. A recent study by economists at Cardiff and Swansea Universities has found that “female directors are associated with less aggressive investment policies, better acquisition decisions, and improved firm performance”, because they are less overconfident.

I’m not sure, though, that this is a perfect solution.

For one thing, if women directors are selected for being atypically overconfident, the problem will re-emerge.

And for another, it’s not just in business that leaders stick with plans in the face of evidence that they are lousy ideas. The same thing can happen in politics, sometimes even with female leaders.