Do Women Lack Confidence?
"Women just aren't as confident as men. That's why they don't advance — the confidence gap is holding them back."
The 'confidence gap' is a real measurement artifact and a largely incorrect interpretation. Women rate their own competence lower than men on self-report scales. But calibration research — comparing self-rating to actual performance — finds women more accurate and men more overconfident. The issue isn't women's confidence; it's men's miscalibration and the social asymmetry in how each is rewarded.
What the data says
-
In a meta-analysis of 26 studies, men overestimate their performance relative to actual outcomes by 30% on average; women underestimate by 11%. Neither is 'accurate confidence.'
Niederle & Vesterlund, QJE (2007) · 2007 · Meta-analysis of confidence calibration studies
-
HP internal study (frequently cited): women applied to jobs only when they met 100% of listed qualifications; men applied at 60%. A more careful follow-up (Mohr, HBR 2014) found men and women both applied when they met ~60%, but women more often reported not applying due to concerns about not meeting the bar.
Mohr, Harvard Business Review (2014) · 2014 · Survey
-
Women's self-rated leadership skill is lower than men's in most studies — but their peer/subordinate ratings are higher than men's on most dimensions.
Zenger Folkman in HBR (2019) · 2019 · 360° review database
-
Barber & Odean: men's overconfidence in investing drives 45% more trading than women, reducing net returns by ~1 pp/year.
Barber & Odean, QJE (2001) · 2001 · 35,000-account analysis
Where it came from
Katty Kay and Claire Shipman's 2014 book *The Confidence Code* popularized the 'confidence gap' framing. It was based on real research but has been partially walked back since — including by critics like Mohr (2014) who re-examined the HP jobs-application data and found the original framing misleading. The more accurate finding is about calibration, not confidence.
What this means
'Women lack confidence' is a framing that puts the deficit on women. The calibration data puts it on men — who are systematically overconfident, which produces worse investing, worse trading, more accidents, and more falling-up promotion patterns. The corrective isn't 'women become more like men'; it's 'evaluators learn to discount overconfidence as a positive signal.'
Frequently asked
Is the confidence gap real?
Yes in self-report surveys — women rate themselves lower on competence. But the gap is about calibration, not confidence: women are closer to accurate, men are systematically overconfident.
Doesn't lacking confidence hold women back?
Yes in settings where evaluators reward expressed confidence over actual performance. The fix is evaluator training, not asking women to be more overconfident.
What about the 'women only apply if 100% qualified' claim?
Mohr's 2014 HBR reexamination of the HP data found the original claim misleading. Men and women applied at similar thresholds; women more often reported hesitation, but actual application rates were closer than the popular claim.
Is being overconfident bad?
Yes, measurably — in investing, medical decisions, project estimation, and negotiation. Overconfidence correlates with worse outcomes in nearly every domain it's been studied. The stereotype treats it as a virtue.
What should women do?
The research doesn't support 'be more confident regardless of evidence' — that's asking women to copy the miscalibration. The supported fix is: collect direct evidence of your own performance, reference it in self-advocacy, and work in environments that reward calibrated over expressed confidence.