Google explicitly rejected merit-based hiring for least-regret hiring by setting the bar much higher than strictly necessary for adequate performance. This gre

Google explicitly rejected merit-based hiring for least-regret hiring by setting the bar much higher than strictly necessary for adequate performance. This greatly reduced the rate of unregretted attrition and, until 2022, layoffs. What this means is that most applications are rejected and so it's very sensitive to unconscious bias or under-recruiting from pools of minority candidates essentially due to network effects. Internal referrals will make up a sizeable percentage of successful hires because they take precedence in the recruiting pipeline (recruiters are still incentivized to maximize their hiring percentage and referrals often have stronger signal to noise on candidate ability than searching through resumes or accepting applications). This effect occurs with more standard hiring practices as well, e.g. "old boys networks", but as the ratio of hires to applicants increases it becomes much more obvious when hiring is biased against minorities and the actual biases can be corrected with policies and procedures. Hence Google's focus on equity in the recruiting pipeline. I believe it was unconscious bias of referrals coming primarily from the same social and educational circles which, under a truly equitable society, would not have been a problem. But it's long been recognized that prestigious schools and social networks are also bereft of minorities from systemic disadvantages and pre-existing network forces. Reaching out to under-sourced schools and pools acts to restore a more uniform selection over actual candidate ability. In short, no one at Google I talked to felt like a "diversity hire". We mostly all felt imposter syndrome due to the least-regret hiring process.