Dunning-Kruger Effect

The cognitive bias identified by psychologists David Dunning and Justin Kruger (1999) in which people with limited competence in a domain dramatically overestimate their ability, because the very skills needed to recognize incompetence are the same skills the incompetent person lacks.

The Dunning-Kruger effect comes from a 1999 study by Cornell psychologists David Dunning and Justin Kruger. They administered tests of logical reasoning, grammar, and humor to students, then asked each student to estimate their own performance. The result: those who scored in the bottom 12% estimated they had scored in the 62nd percentile on average. Not just slightly overconfident, massively, systematically overconfident.

The mechanism: competence and the metacognitive ability to recognize competence are not separate things, they're the same cognitive toolkit. Someone who cannot solve logic problems also cannot recognize when they are failing at logic problems. They have no reference point for 'correct reasoning.' The expert's curse inverts this: highly competent people often underestimate themselves because they know how hard the problem is and assume others find it equally hard.

The double peak: the full Dunning-Kruger curve shows not just the competence valley but also a second phenomenon, experts underestimate their relative advantage. Those in the top 5% typically think they're around the 70th percentile. Both the worst and the best systematically misassess themselves; the mediocre are most accurately calibrated.

Political application: Dunning-Kruger provides a partial explanation for why confident incompetence rises in democratic politics. Voters cannot always distinguish competent from incompetent politicians. High-confidence-low-competence individuals perform well in environments rewarding charisma and assertiveness over measured accuracy (political campaigns, media). As the Game Theory series notes: 'Often the people in power are stupid, they don't know they're stupid, they are confident, and they do stupid things.'

Frequently asked questions

What is the Dunning-Kruger effect in simple terms?

People who are incompetent at something tend to wildly overestimate how competent they are. The reason: the mental skills needed to recognize you're bad at something are the same skills you'd need to be good at it. If you lack those skills, you also lack the ability to see that you lack them. People who scored in the bottom 12% of logic tests estimated they were in the 62nd percentile.

Does the Dunning-Kruger effect mean confident people are incompetent?

Not exactly. It means incompetent people tend to be overconfident, but competent people tend to be underconfident. The full curve: the worst overestimate dramatically, the mediocre are fairly accurate, and the best slightly underestimate themselves (they know how hard the task is and assume others find it equally hard). Confidence is not a reliable signal of competence in either direction.

← Back to Glossary