Harvard discrimination lawsuit: data show penalization of Asian-Americans

I really don't see it as Harvard enforcing racial quotas but rather looking at the whole student not just academics and standardized test scores. I applaud them for attempting to create a diverse and varied student body. Do you have a better alternative or would you select a student body just by GPA and standardized tests?

An anecdote about the evolving value of higher education:

I received a resume recently for a junior software engineer role. Under the 'Education' section was a paragraph that listed a prestigious university and entrance year, but under it stated something like, "I was accepted to this university, but rather than attend and go over $100,000 in debt, I taught myself computer science and coding for 3 years with online tutorials. Here are the projects I did over that time," with links to a github repo showing the projects from his self study.

I gave him a phone screen, and while he wasn't a fit, I did not exclude him because of the nature of his education.

I think more hiring managers in companies that are struggling to find talent will begin to accept this narrative about education. The signal to me (if it checked out) was his acceptance to a prestigious school, and his lack of attendance was not very important to me, provided he could do the work.

Very easy. Admissions should be race blind and name blind. Everything else, including socioeconomic background, can be included in the admissions packet.

It seems like there are two conflicting goals:

- don’t explicitly consider race, gender, etc., while you make a decision about a cohort of admitted applicants. That is, only use other “performance” criteria (grades, essay, recommendations, etc).

- ensure that post facto, the observed distribution of demographic properties of your cohort of admitted students fairly reflects the natural distribution, within a societally acceptable margin of error.

The problem is that these two goals are in conflict for at least two reasons, one big and one small.

The small reason is that a person’s most salient “performance” metrics might involve direct relationship to their role in a religious community, race-related organization, etc., such that it is impossible to treat it like a truly demographic-blind process.

The big reason is that for whatever performance criteria you pick, like grades, test scores, essays, community involvement, etc., there is endemic and generational discrimination and privilege issues that can mean even if you truly did not want to make decisions based on race, religion, nationality, etc., you inadvertently did just by using proxies for societal scale discrimination or disadvantage in those categories.

So in the end, you can see why a big organization or corporation is terrified and would rather use quotas to explicitly control the relative ratios and acceptance ratios on a category-by-category basis.

I’m not trying to defend it. I think it’s a bad thing.

But what do we do? It seems like a lose-lose.

- If you ensure perfect decision blindness to race, religion, nationality, etc., you’ll end up admitting a cohort of students that greatly favors basically wealthy white people, because whatever proxies you do use will encode longterm privilege deeply enmeshed in them, way beyond whatever we are consciously aware of.

- If you do explicitly consider race, religion, nationality, etc., to try to manually control rates of admission per each possible category, you can ensure the admitted cohort has the overall demographic statistics you want, but only by explicitly making race, religion, etc., as major driving factors that are explicitly and consciously weighted into the admission decision.

So, you disagree with the allegation that the admissions office was biased against Asian-Americans in the personality test?