<p>It’s appropriate and accurate for me to say “some.” </p>
<p>It’s just math.</p>
<p>The 90-10% difference doesn’t support your claim that MOST or ALL URMs benefit. That figure only holds (with the caveat about methodology cited before) for the specific number of applicants in 2005 who applied with exactly a 1250 SAT and 3.2 GPA. Do you know what number that is? Relative to the overall pool of applicants? I don’t have that information with me, but I am certain that it has to be small given the profile of applicants to U-M and the relatively small number of URMs in the pool. It would be spurious to say that the 90-10% split applied to ALL URM applicants. I also think it would be very questionable to claim that the 90-10 split applied to MOST URM applicants.</p>
<p>The basic point is this: the advantage that a minority candidate got from the former SI varies depending on his or her overall profile. For the most qualified candidates, it mattered not at all. For some number of marginal candidates (I mean marginal for admits thresholds at the U), it mattered enormously. And for very poor students, it also didn’t matter. 20 points isn’t going to help a bottom-basement SI. </p>
<p>Put another way: It ONLY matters for that number of candidates whose SI without the 20 points would have fallen below the admit threshold. That is, if U-M was admitting at SI of 125 or above, it mattered for all the URMs with a 144-125 SI. Because without the 20 points they would not have gotten in. Above 144, it didn’t matter. Below 125, it didn’t matter. What portion of the URM applicant pool fell between 144 and 125? All? Most? I don’t know, so I’m saying “some.” I think this is justified.</p>