Glenn Loury had a woman on who is the head of the Chinese-American something-or-other challenging the blatant discrimination against Asians in placement in the seven NYC advanced high schools. At present, 97% of the admissions have been on the basis of a single test, administered yearly. The other 3% are students who were very near the cutoff, but can demonstrate some sort of disadvantage. Asians currently have about 50% of the places, but under pressure to become more demographically similar to New York as a whole, deBlasio wants to move to a different balance, which would have about 25% Asians, and more Blacks in particular.
She is absolutely right, but people just don't like reality. She made an excellent point when Loury, attempting to clarify and summarise, used the words "under-represented" and "over-represented." She countered that these children aren't "representing" anything or anyone but themselves. That is frankly, a very American or even simply Western idea we might be glad that people from the East have figured out and adopted just fine. Yet that is not the prevailing thinking in education, nor in liberal advocacy groups in general at present.
However, I wonder how much it matters. I don't think schools matter that much in student outcomes. When we look at school district outcomes not in terms of final product, but in terms of difference between kindergarten and graduation testing, they aren't that different, and the differences can usually be tied pretty clearly to demographic changes during those 13 years. Hanover High School likes to think of itself as the best public school in the state on the basis of its highest test scores every year, but its two main industries are Dartmouth College and world-class Dartmouth Hitchcock Hospital distributed over a small population of a few towns. Those kids come into kindergarten with the best scores in the state as well.
Schools that are not safe matter, though measuring even that has the confounding factors that the surrounding neighborhoods, and the genetics of the people in those neighborhoods is not the same either. Let us grant that there is likely some depression of the education of some kids who might otherwise have succeeded, even though I don't know of studies that demonstrate it. I also admit that most studies of schools are not measuring the effect on the brightest students, but on the population as a whole. There might be some real effect of putting the brightest kids together. I doubt it. The advantages are going to be the same ones that the kid entered with, plus whatever the prestige factor is.
However, the more that colleges don't screen for testing and move toward the much more discriminatory holistic admissions, the more that the selective places will secretly rely on the remaining places that do test, regarding mere admission there as their clue. Graduate programs forced to stop looking at GREs will just look at where you got into college instead. Fewer diamonds in the rough going forward.