Are experts overconfident?: An interdisciplinary review

Abstract

Are experts overconfident? Some research finds experts are plagued by overconfidence whereas others conclude that they are underconfident. We reviewed the literature, taking an interdisciplinary approach, to answer this question. In doing so, we assessed whether there were theoretical differences in how overconfidence and expertise were conceptualized across the literature. For overconfidence, there are three distinct conceptualizations with a fourth captured by economic models. People can be overconfident because they forecast a narrow range of possibilities that the truth frequently falls outside of (i.e. overprecision), overrate their judgments as correct (i.e. overestimation), or overbelieve they outperform their peers (i.e. overplacement). In economic models, overconfidence is at times assessed by the likelihood of engaging in unwise behavior and making inaccurate predictions. Likewise, there are divergent definitions of expertise. People can be identified as experts because of their experience (i.e. time on task), because of enhanced performance on knowledge tests, or because of their job title or professional degree. These conceptual and theoretical inconsistencies are crucial in answering whether experts are overconfident. For overprecision, experts are overprecise across all theoretical definitions of expertise. However, this consistency is likely because overprecision is so robust across people regardless of whether they are experts. For overestimation, we found experts of experience and experts of title typically overestimated, with mixed results for the experts of knowledge. Studies of overplacement are limited, often defining experts via knowledge, but find that they tend to underplace their abilities. Last, when overconfidence is assessed via economic models, experts display overconfidence.