Flying high, falling far

Written by Saurabh Bhargava | Updated: Sep 4 2009, 03:53am hrs
A rare point of agreement amongst major media, in recent months, has been to implicate overconfidence as the root cause for last years economic crisis. The source of such overconfidence depends on whom you ask. An early and angry chorus denounced the arrogance of Wallstreet CEOs. Others took partisan aim at current or past administrations. Jon Stewart censured Jim Cramer and CNBC. And some found fault in the overly cheery forecasts of economists.

While adjudicating the relative merits of these indictments is no easy task, the broader case against overconfidence seems more transparent. Most people are convinced that they are better than average in most domains. Whether it is the likelihood of securing high returns in the stock market, the susceptibility to sickness, or even the ability to drive, people tend to view themselves with unwarranted generosity.

One can think of overconfidence as the tendency to overestimate absolute or relative ability, as well as the tendency to believe that ones knowledge is more accurate than it really is. Psychologists have documented such overconfidence for decades, usually by quizzing subjects in a laboratory and then making predictions regarding accuracy. This research has erected an impressive foundation of human immodesty.

Why is overconfidence so embedded in our constitution and why arent we able to learn from past mistakes First, it may be that our overconfidence, like many foibles, conveyed an evolutionary benefit that is no longer quite as advantageous as it once was. Belief in ones competence conceivably provided defence against adversity that outweighed the costs of lessened objectivity.

Second, overconfidence, like other self-enhancing biases, may arise from errors that we make in attribution. That is, we generously attribute favourable outcomes to our skill, while for less favourable outcomes, we blame contingencies. Meanwhile, when evaluating others, we are not quite as charitable. The next time you evaluate stock market performance, note whether the invocation of aptitude or luck hinges on whether the returns are positive and whether or not they are yours.

Another possibility is that people tend to infer more control over situations than they actually have and that such illusions of control lead to overconfidence. One can find theatrical demonstrations of this at the nearest casino. Gamblers throw dice with gusto when seeking a pair of sixes and with soft finesse when the aspiration is more modest. One can detect less fragrant evidence from superstitious athletes who insist on not changing or showering so as to sustain lengthy winning streaks.

A final explanation implies shortcomings in the way in which we evaluate evidence and make inferences. Suppose we perceive uncertain outcomes as closer to average than they really are because the average represents a plausible best guess. Because we have better information about ourselves than others, we will then tend to evaluate others more regressively than we view ourselves. An implication is that individuals will be overconfident about relative success in easy tasks where performance is high, but will exhibit underconfidence in the face of more difficult tasks. True to form, surveys show that beginning drivers actually believe they are worse, not better, than the average novice driver.

Beyond increasing gambling costs and the malodorousness of sporting events, is overconfidence all that bad Delusions that your voice sounds slightly more like Bono than it is seem harmless enough. So long as you dont sell the house to buy recording time in the studio, perhaps ignorance is really bliss, and some self-deception is constructive. Surely, overconfidence disappears when the stakes are high enough and the sufficiently experienced are calling the shots Unfortunately, the answer may be no.

Consider the popular claim, first voiced by Alan Greenspan, that the stock market is subject to irrational exuberance. One strain of exuberance is the pervasive belief of most investors that they are more able than their median counterpart. A consequence of this overconfidence is that investors tend to trade more frequently than is optimal and this excessive trading leads to lower returns on the order of 2-3%. Incidentally, in financial decisions, men tend to be more overconfident than women. One study of a large discount brokerage determined that men traded 50% more often than women creating a 1% gender gap in returns.

Other examples of the perniciousness of overconfidence also abound. A colleague and I are currently exploring whether overconfidence with respect to ones ability to work late into life, as well as to avoid negative health shocks, contributes to the worrisome lack of savings amongst middle-age America.

Negotiation scholars argue that overconfidence in valuations amongst bargainers widens buy-sell spreads, increases the rate of impasse, and results in higher rates of litigation.

Lest you begin to take solace in your own professional credentials, note that the effects of overconfidence are not confined to weekend gamblers and neophyte stock pickers. In one recent study, Berkeley and UCLA researchers characterised CEO overconfidence by examining the degree of personal exposure to firm-specific risk and by looking at media portrayals. They then found that overconfident executives are 65% more likely to engage in value-destroying acquisitions.

Moreover, historians and political scientists have long cited overconfidence of political and military rulers as an important cause of international conflict. When former CIA Director, George Tenet, allegedly offered a slam-dunk assurance regarding the discovery of WMDs in Iraq, he provided a vivid and recent illustration of the possible geopolitical consequences of overestimating the precision of ones beliefs.

The emerging picture is not a pretty one. Overconfidence seems far from innocuous and, depending on the context, may be downright catastrophic. It afflicts novices, experts and possibly even shapes the decisions of nations. It appears largely immune to the usually corrective nudge of experience.

It leads to a variety of ills from poor investment returns and lack of preparedness for earthquakes to low take-up of preventive medicines, even to war. So what can be done

Scholars and practitioners offer a menu of strategies to combat egocentric biases such as overconfidence. These include actively empathising with competing points of view, pursuing hypothesis disconfirming, rather than confirming, evidence, enlisting a neutral third party, and developing clear and transparent ex-ante metrics for measuring success. But maybe it comes down to this. Identify all of the domains along which you are just average or even fall below average. And then take heart in the realisation that at least along one critical dimension, self-awareness, you are now undoubtedly better than average.

The author is adjunct faculty in economics at the University of Chicago, Booth School of Business