When overconfidence is revealed to others: Testing the status-enhancement theory of overconfidence

Functionalist theories of status suggest that groups would penalize confident individuals with lower status if they were to discover that the individuals’ confidence is unwarranted. Groups accord lower social status to individuals who hinder the group’s success (e.g., Blau, 1964, Ridgeway and Diekema, 1989). Groups might thus penalize overconfident individuals because overconfidence jeopardizes task performance (Barber and Odean, 2000, Klein and Kunda, 1994, Metcalfe, 1998, Paese and Kinnaly, 1993, Vancouver et al., 2002).

Moreover, groups might view individuals with unwarranted confidence as more selfish and less committed to the group’s success, and thereby accord them lower status (Willer, 2009). In laboratory experiments, groups ostracized individuals who claimed more status than the group believed them to deserve and paid them less for their work (Anderson et al., 2008, Anderson et al., 2006). Taken together, these findings suggest groups might penalize overconfident members with lower status. In other words, groups might accord overconfident individuals a lower level of status than that accorded to others with the same level of competence but more accurate self-perceptions of ability.

The above arguments notwithstanding, we propose that groups will not penalize overconfident individuals, even after discovering that those individuals’ confidence is unwarranted. Indirect evidence cited below suggests that high levels of confidence might create positive peer-perceptions that remain even after actual task performance is revealed to others.

These resilient positive peer-perceptions may take two forms. First, confident individuals might appear more socially skilled to others, even when their confidence is unwarranted by actual task abilities. Groups accord higher status to those with perceived social skills (Bass, 2008, Lord et al., 1986, Stogdill, 1948). Social skills contribute to a group by helping to coordinate other members’ activities, solve conflicts, and motivate others while maintaining cohesion within the group (Bass, 2008, Fragale, 2006, Mann, 1959, Van Vugt, 2006). Individuals who are confident may appear socially skilled by acting more engaged, speaking more often, and participating actively. Communication is a key aspect of social skill (Hall, 1979, Riggio, 1986) and individuals who communicate more are often seen as more skilled (Breland and Jones, 1984, Littlepage and Mueller, 1997, Littlepage et al., 1995). Consequently, these individuals receive attributions of greater leadership ability (Mullen et al., 1989, Sorrentino and Boutillier, 1975, Sorrentino and Field, 1986).

In addition to increasing quantity of communication, confidence may also reduce anxiety about participating in the task. This may result in more fluid, clear, concise speech and smoother social interactions. Past studies have found that lower anxiety relates to higher performance on a variety of tasks, from academic tests to interpersonal interactions (Brooks and Schweitzer, 2001, Glass et al., 1982, Littlepage et al., 1991, Osborne, 2001, Plaks and Stecher, 2007, Steele, 1997). With less anxiety, confident individuals, even when their confidence is unjustified by actual task ability, may be more articulate and attentive to others. The combination of more frequent participation and more fluid social interaction may lead to impressions that confident individuals possess superior social skill.

Second, high levels of confidence might create positive perceptions of task ability that persist even in the face of contradicting evidence. First impressions exert lasting influence on interpersonal judgments (Benassi, 1982, Jones et al., 1968, McAndrew, 1981, Mussweiler, 2003, Steiner and Rain, 1989, Zenker et al., 1982). Impressions of task ability might not completely adjust to account for the objective feedback provided. Therefore, even when objective information on task performance contradicts initial impressions, group members may perceive confident individuals to possess higher task ability.

If confidence creates persistent peer impressions of social skill or task ability, groups may not penalize confident individuals with lower status, even after discovering these individuals’ confidence is unjustified by their actual task skills. Instead, upon discovering that individuals are overconfident in their task abilities, groups might simply accord those individuals the status they appear to deserve based on their true task performance or perceived social skills.

Central to the current research is whether the status benefits of overconfidence outweigh the potential status costs if one’s overconfidence is revealed to others. If so, this helps further support the status enhancement account of overconfidence. One way to test whether the status benefits outweigh the costs is to examine whether overconfidence yields a “net” positive status outcome for the individual on average. To illustrate in a simplistic way, if being overconfident led an individual to gain an “extra” 3 status points when her overconfidence was unknown to others, but to then lose 5 status points when her actual task performance was revealed, overconfidence could be considered to have a negative net value. Averaging across the two situations, overconfidence would have a net status value of -1. However, if being overconfident led an individual to gain 3 status points when it went undetected, but to then lose only 1 status point when his actual task performance was revealed, overconfidence could be considered to have a net status value of +1 (i.e., a positive net value). We assessed net status outcomes in the current studies and predicted that confidence would yield net status benefits, even when the confidence was unjustified by actual ability.