Gender Gap by Board

Gap estimates are doubly noisy — each gap combines sampling error from two subgroup proportions. At board level, subgroup counts are generally large enough for stable estimates, but interpret gaps smaller than ~5 pp with caution. See measurement reliability for details.