Indeed, at scale, a company that is significantly unrepresentative of the population from which it draws its employees surely has something wrong with it. That’

Indeed, at scale, a company that is significantly unrepresentative of the population from which it draws its employees surely has something wrong with it. That’s basically what these studies show (flaws notwithstanding). But that’s got nothing to do with individual teams (unless they’re huge) which are far too small to exhibit population-level statistical behaviour. I see this claim repeated often but the sources always end up being these flimsy studies of company-wide public company reporting correlated with market performance.