Ok, let's do some nice, calm soothing math for a second. Per the OP previously 'diversity' candidates were 1% of all applicants. I'll assume the null hypothesi

Ok, let's do some nice, calm soothing math for a second. Per the OP previously 'diversity' candidates were 1% of all applicants. I'll assume the null hypothesis here and then also assume that they were 1% of the interviewees too. Lets assume that diversity interviewees are turned into hires at some factor Y. I'll make no assumptions on if that is different than from non-diversity interviewees. Now, with the new policy, there is a 15x increase in the probability of turning diversity applicants into interviewees. However, nothing has been done to change Y, the factor at which diversity interviewees are turned into candidates. They are explicitly stating that they are not changing Y. So that then means that diversity interviewees are now less likely to move on past the interview stage. Based on the numbers, they then need to interview at 15x the rate as previously to be turned into hires. Please, correct me if I am wrong here, but this seems to hurt diversity interviewees. I see it as taking up 15x the time, rejecting at a 15x rate, and eliciting these real human people to become stats in some database that the policy makers can show off to some other boss without any compensation.