It Pays to Be Overconfident, Even When You Have No Idea What You’re Doing

Photo: Jason Squires/WireImage

In his New York Times column earlier this week, David Brooks, responding to an essay in The Atlantic about how women have less confidence than men, wrote that "recent psychological research … suggests that overconfidence is our main cognitive problem, not the reverse.”

It’s certainly easy to come up with examples of overconfidence getting us into trouble — the Iraq War, the financial meltdown, that guy who challenged a heavyweight boxing champion to a fight last week — but overconfidence may actually be beneficial, at least for the person with the big head.

Consider Kanye West, one of the greatest bloviators of all time. Here is a partial list of people Kanye has compared himself to: Michael Jackson, Picasso, Beethoven, Steve Jobs, Walt Disney, and of course Jesus, who presumably died so that Yeezus could live. Even in an industry built on braggadocio, his crowing seems excessive. But it may also be the key to the spell he casts, even over the haters: According to new research, overconfidence increases one’s status even when it’s been exposed as overconfidence.

Of course, Kanye comes from a long line of overconfident beings — the whole Homo sapiens species. Psychologists argue that we have evolved to overestimate our abilities, to see ourselves as better than our peers, and to be overly certain about our judgments. People regularly show overconfidence about their athletic ability, the quality of their marriages, and the promise of their academic research. In one study, 93 percent of Americans declared themselves better-than-average drivers.

So we all live in Lake Wobegon. But whether or not we consider that a character flaw, it may be an evolutionary asset. In recent years, some researchers have been pushing the idea that overconfidence, and self-deception in general, evolved to fool others, a notion first proposed by the biologist Robert Trivers and developed further in a 2011 book. We deceive ourselves about our superiority so that we may better deceive our potential competitors, collaborators, benefactors, and mates. To be a good salesman, you have to buy your own pitch.

In 2012, Cameron Anderson at Berkeley and his colleagues published a paper supporting this idea, showing that overconfidence increases one’s status. Subjects who overestimated their abilities at group tasks were more respected and influential in the group. It turns out, we tend to (over)use confidence as a useful proxy for competence — if you speak firmly, it sounds like you know what you’re talking about. People who showed more confidence, regardless of their actual ability, were judged to be more capable and accorded more regard by their peers.          

But a new paper has even more striking findings, with implications for those who overestimate their skills so grievously that you might expect to see backlash (see Trump, Donald). Jessica Kennedy of Vanderbilt, working with Anderson and Don Moore of Berkeley, showed subjects videos of actors playing subjects from a previous experiment. In the clips, the actors displayed either a medium or high amount of confidence during a group discussion, and subjects rated those displaying greater confidence as having more status (respect and influence) in the group. Then they were told the targets’ purported scores on a related task — they were in either the 47th or 91st percentile — and asked to rate them again. Even after learning the scores, subjects saw average performers as having higher status when they had lots of confidence than when their confidence matched their abilities.

The paper also reveals two factors that may buoy the status of the obviously overconfident against the weight of censure: greater confidence leads to greater peer-rated social skill and greater peer-rated task ability, regardless of actual ability. The researchers suspect that confidence increases leadership-like behavior, such as talkativeness and active engagement, and also reduces anxiety, which allows for more fluid interaction, and that these behaviors may make one seem more socially skilled. As for the effect of confidence on perceived ability even after actual ability has been reported, the authors note the lasting power of first impressions have been long known to disproportionately affect our judgments of others. All of this suggests that even when we’re unmasked as less skilled than our self-assured manner would suggest, there are ancillary social benefits to overconfidence.

Maybe this is how pundits (and Times columnists) maintain their audience, and why political candidates feel free to make undeliverable campaign pledges: There may simply be insufficient downside to their overpromising. The logic applies even if your leadership role is on a slightly less grand scale than the leader of the free world’s, according to a paper Connson Locke of the London School of Economics and Anderson have submitted for publication. In one study, subjects discussed two hypothetical job candidates with someone playing their superior. The person playing their superior always argued for the candidate who was clearly (to the subject) weaker. When he acted more confidently, the superior was rated as more capable, even though he was expressing his confidence in favor of a clearly weaker candidate. This confidence, even in the wrong choice, made subjects more likely to defer to him and select the less optimal job applicant. So it’s easy to deride overconfidence as a big societal problem (which it is), but we shouldn’t overlook the reason it’s so prevalent: It yields real benefits for the people who exhibit it.

Some of these findings about the power of confidence could help explain the gender gap in leadership positions at big firms, since there are studies showing that men tend to puff their egos up more than women (a fact lamented in the Atlantic article that spawned Brooks’s column, as well as the book it’s adapted from). This self-deception likely enhances their status.

Would Anderson advise people to be overconfident? “God, that’s a hard one,” he says. “The truth is, I keep showing that it benefits people, but the reason I started studying this is that I was tearing my hair out watching overconfident people get ahead.”