In strictest overconfidence
By David Stauffer

IN a look at possible explanations for discounter Kmart’s loss of market share under the leadership of CEO Joseph Antonini, Wall Street Journal reported that "attitude may have made a bigger difference than strategy.... Antonini didn’t think others could tell him much about the business... he bristled at criticism and... didn’t do much hiring of managers from outside the company who might challenge him."

If this assessment is correct, Antonini was the victim of overconfidence — the understandable and all-too-human tendency to accept our own experiences and beliefs as reliable truths. We humans, it seems, are almost irresistibly motivated to use even the flimsiest evidence as the basis for unshakable conviction.

Further, we are convinced that we have a solid basis for our conviction, when in fact the vast majority of us are stunningly over confident. And even when we are incontrovertibly proven wrong by events or the weight of other opinions, we are inclined to go in a direction that compounds rather than alleviates the damage resulting from our original position.

These powerful tendencies are most vexing to people who are frequently called on to make decisions and those whose careers have been marked by steady advancement toward the top ("I wouldn’t have been promoted for being wrong, would I?"). Fortunately, if you can accept these tendencies as part of the package that comes with being human, you can neutralise their ill effects and sometimes even turn them to your advantage — as superstar leaders Jack Welch of General Electric and Andy Grove of Intel have done.

Decision research shows that we groundlessly bolster belief in the extent of our knowledge. Among the more prominent ways we do this by:

l We remember little of what we experience. University of Bamberg psychology professor Dietrich Dormer, in his book The Logic of Failure observes, "Human memory may have a very large capacity, but its ‘inflow capacity’ is rather small. What we perceive at any given moment may be rich in content, colourful and clear in its contours. The moment we close our eyes, however, a great deal of that richness instantly disappears — unclear and pale outlines remain".

l We remember selectively. In The Challenger Launch Decision, Boston College sociology professor Diane Vaughan observes that even highly skilled and experienced professionals have been shown to selectively accept or dismiss new facts and evidence to fit their existing world view. The result: decisions based on preconceptions rather than current circumstances.

l We remember prejudicially. Eliot R. Smith, a Purdue University psychology professor, studies the effects of exemplars — mental pictures of, say, a least-liked school teacher — that we unknowingly carry with us and apply to others. The danger? If a new acquaintance, perhaps a prospective new hire, reminds you of one of your exemplars even "below the level of consciousness," then you will judge the new person by what you think of the exemplar. Smith relates findings of an experiment in which participants had an encounter with someone with a particular hairstyle who was mean to them. Later, when presented with a new, not-unfriendly person with the same hairstyle, the participants judged that person to be unfriendly. "It doesn’t make sense," Smith notes, "but it still has an effect".

lWe change what we expected to fit with what actually happened. Vaughan observes that evidence of the design flaw that led to the 1986 Challenger tragedy was first seen in 1977. But the "deviancy" from expected performance was "normalised" — engineers found an explanation for it. Subsequent manifestations of the same problem over the years were treated the same way: Deviances from past performance were in fact predicted.

We toss these and other monkey wrenches into what we swear is "knowledge", giving us a gigantic case of overconfidence, according to J. Edward Russo, a Cornell University professor of management who conducts executive seminars on managerial decision-making. He and Wharton School research director Paul J.H. Schoemaker have administered "overconfidence quizzes" to thousands of managers around the world, consistently finding that the managers grossly overestimate what they claim to know.

Russo says that the attitude of one bank’s chief loan officer is typical. "He claimed that he and the loan officers who worked for him would score very high on a quiz which he himself assessed as being germane to his industry and not particularly difficult. ‘No one is more realistic than a banker,’ he said. He then took the test — and failed miserably. He had his loan officers take the same test. Every one of them flunked".

In a 1992 Sloan Management Review article, Russo and Schoemaker explained that their overconfidence quiz "measures something called metaknowledge: an appreciation of what we do know and what we do not know.... We draw on our metaknowledge when we conclude that we have enough information and are ready to make a decision now.... No group of managers we tested ever exhibited adequate metaknowledge; every group believed it knew more than it did about its industry or company. Of the 2,000-plus individuals to whom we have given a 10-question quiz... fewer than 1 per cent were not overconfident".

One group that appears to have fallen victim to a very public case of costly overconfidence is the General Motors corporate design staff who redesigned the 1991 Chevrolet Caprice. Numerous press accounts noted the staff’s rejection of negative comments on the prototype, such as this one in the Washington Post. "The design staff... had a reputation for being resentful of marketers, engineers and other corporate outsiders who had the gumption to tell them what was and was not attractive". The article goes on to report that when a focus group also voiced its strong objections to the proposed new Caprice exterior, design staff VP Charles M. Jordan "didn’t like the clinic’s reaction. ‘We were excited about the design,’ he said. ‘We decided not to do anything about it. We believed in the design.... All the car guys liked the design".

Caprice sales for 1991 amounted to half the anticipated volume, and in 1995, the model was discontinued.

But the Caprice saga is distinguished in the annals of overconfidence only by its very public unfolding. "This is the stuff of human nature," Russo observes. "It is apparently worldwide. We’ve tested in Europe and Asia as well as North America. Asians, admired in many ways for their business acumen, are, if anything, even more overconfident than Americans. It’s not culturally specific — either in terms of societal or organisation culture".

It’s difficult, not to mention personally risky, for a subordinate to try to get a message of reality through to the overconfident-and-proud-of-it boss. "There’s no way to convince that boss to hear what he doesn’t want to hear," Russo laments. "It would be like the Hollywood Mogul who was reputed to have said ‘I don’t want a bunch of yes-men around me; anyone can say no if he’s willing to lose his job".

Nonetheless, Russo says, "Overconfidence isn’t all bad. It fills our need to believe in our abilities. It can contribute to a palpable optimism, which has motivational value. And an argument could be made that risky projects are undertaken when some key people have an unrealistic belief in their chances of success".

We’ve seen that we can easily be led to errant decision-making by believing we possess a level of knowledge about our industry or field of expertise that we, in fact, overrate by 100 per cent or more.

Now for the really bad news: Human nature doesn’t just lead us to mistaken judgments by routes such as those discussed above; it invites us to follow our initial misjudgment with responses that, at best, do nothing to correct our course or, at worst, lead to tragedy. Thus Vaughan recognises, "The O-ring failure that destroyed the Challenger was preceded by a year-long series of sporadic O-ring problems".

A series of decisions with increasingly poor outcomes is called a "doom loop" by Eileen C. Shapiro, president of the Cambridge, Massachusetts, consulting group Hillcrest Group Inc., in her forthcoming book The Seven Deadly Sins of Business. In an interview, Shapiro said, "Many executives get caught in this downward spiral, where unquestioned and incorrect beliefs lead them to compound rather than eliminate their past mistakes. Look at what happened when consumers began to turn away from microwave ovens that were progressively loaded with more and more features, options and complexity. The response by some manufacturers was not to simplify things, but to offer cooking classes, more elaborate instructions, and — incredibly — still more features".

"The single most important step you can take is to allow your beliefs to be called beliefs and not truths," says Shapiro. Russo agrees, adding, "You have to have that crack in the door: the boss recognising at least the possibility of his fallibility. It involves an absence of self-delusion that is itself closely related to overconfidence — the leader who isn’t afraid to admit what he doesn’t know. Jack Welch is reportedly such a person. If so, the phenomenal value he has added for GE probably results at least in part from his willingness to acknowledge he may not have all the answers".

Russo also reports that some executives may overcome overconfidence with awareness alone. He cites a study in which half of the participants in a negotiation exercise were warned about overconfidence: "Compared to the unwarped group," he says, "those forewarned were 30 per cent more likely to reach a negotiated agreement instead of having to turn to costly arbitration, and they achieved net dollar benefits that were 70 per cent higher".

Similarly, "good managers devise their own solutions to the problems of overconfidence," Russo continues. "That head loan officer who had been so overconfident of his knowledge devised a ‘competitor alert’ file for his term of loan officers — they were required to contribute to it and read everyone’s compiled submissions. Only three weeks after initiating the file, one of the officers was alerted to the possible defection of a major client to another bank. He contacted the client and convinced him not to switch, saving $ 160,000 in annual revenue".

"As strange as it may seem," says Alcock, "experience is often a poor guide to reality. Critical thinking helps us question our experience and avoid being too readily led to believe what is not so. In effect, we come to a conclusion and they say, ‘Hold on, how do I know this?’ We suspend judgment. We can’t apply this to everything we do — for example, standing at the supermarket dairy case and thinking, ‘This looks like a quart of milk, but is it really?’ But we can apply it in the case of decisions that are outside our areas of expertise".

Russo reflects the consensus of opinion in concluding that "group judgements on average, (are) better than individual judgements" in decision exercises. But he and other experts caution that listening to others involves its own pitfalls. Foremost among these is the natural tendency to seek out people (and literature, anecdotes, etc.) who agree with us. In his book Judgment in Managerial Decision Making, Kellogg School professor Max H. Bazerman mentions the special susceptibility of corporate leaders to this tendency: "Prior successes often reinforce this behaviour. Almost by definition, a track record of success means there is lots of evidence confirming what one believes".

Alcock warns of another pitfall: "You have to watch out for those people making the same mistakes you might be making. The corporate leader, above all, must be wary; he is likely to be hearing from people who want to aim favour and may do so by not raising problems or objections to the boss’s idea".

The danger of such behaviour is particularly acute in the presence of a forceful and strong-willed boss. The tough-guy reputation of former Eastern Airlines CEO Frank Lorenzo, for example, is said to have so intimidated top executives that they dared not defy him. A former colleague says lieutenants would strain to catch clues of Lorenzo’s position on an issue and then rush to get on board with the boss.

There are several ways in which you might be able to gauge and adjust for the yes-man factor. First, lay out problems and perhaps alternative responses, but don’t state an opinion until others have stated theirs. Second, state an opinion and gather comments from aides, then drop the issue for a week or two. Bring it up again by taking an opposite stance, and see who switches with you. Third, be truly receptive to dissenting opinion. Go beyond lip service; thank subordinates immediately and publicly for stating their objections. And fourth, formalise dissent and counter-argument; this measure is unanimously advocated by the experts. Russo, for example, champions "counter-argumentation, which has proven so valuable in several studies that it’s probably smart to insist that one or two people take a devil’s advocate role, even when your team’s opinion on an issue is unanimous.... Other studies have found that, when listing pros and cons, the cons do the most good in countering overconfidence".

Questioning yourself enough to avoid overconfidence — but not so much as to cause needless self-doubt — is a tricky balancing act to maintain, especially since we are rarely willing to admit that we could be the victims of our own hubris. But if the alternative is to realise at some point in the future that you could have avoided a mistake, it suddenly seems easier to take a deep breath and ask yourself: How do I know?

(Courtesy Span)