Hacker News Reader

Comment by nonethewiser | Hacker News Reader

I dont think it should even be legal to ask for the applicant’s race nor track their students race. I think the most appropriate response to “How many black people are at your college?” would be “I dont know we dont count them.”

View on HN

> I dont think it should even be legal to ask for the applicant’s race nor track their students race. I think the most appropriate response to “How many black people are at your college?” would be “I dont know we dont count them.”

So if a college is 100% white in a predominately black area, would you not consider that to be somewhat suspicious? What about the reverse? What if a college instituted a test that has a high chance of weeding out people of a certain race incidentally? Is that institutional racism, or is it mere chance?

plurium interrogationum - why are you doing it?

Because we have a massive ongoing and structural problem with racism in America, we need to be doing something productive about it, and there's no meaningful way to do that if we have no idea how many people of various marginalized groups are in places they're traditionally excluded from.

> Because we have a massive ongoing and structural problem with racism in America, we need to be doing something productive about it

Well getting rid of institutional racism from affirmative action is a good start

> So if a college is 100% white in a predominately black area, would you not consider that to be somewhat suspicious?

Assuming that was due to discrimination that would be bad. Not tracking race does not beget racism. You can oppose both. Furthermore you cant prove racism or bias from an uneven distribution of race. If we stopped tracking race maybe we could avoid this logical fallacy and focus our efforts on real racism. Like policies that explicitly authorize discrimination by race.

> Assuming that was due to discrimination that would be bad.

But if it was accidental it wouldn’t be bad? I think the OPs point is that we can’t always know the second order effects of our decisions (hypothetical ex: requiring high test scores from a standardized test that native English speakers do better on) without keeping track of the outcomes.

Exactly. There are multiple cases where things a university does inadvertently favors one subgroup or another. We know for example the SATs favor Asian-Americans. So does having standardized testing as part of your admissions process count as affirmative action? How many layers of indirection do you need in order for it to be acceptable?

It's ultimately not an easy thing to determine at all and trying to just hide the racial statistics of universities is just admitting it isn't a problem as long as it isn't seen.

>What if a college instituted a test that has a high chance of weeding out people of a certain race incidentally

Like the "personality score" which somehow curiously scored Asians lower, without even talking to them in person ?

Don't you worry, the universities will make sure that their decisions are so opaque that they won't ever be caught red handed again

Or what about SATs, which favor Asian-Americans over every other race? Do you believe that the SATs function as a form of affirmative action and should be removed? Does removing them count as a form of affirmative action because it equalizes the field for other individuals?

With the SAT, there is nothing that intentionally restricts other people from performing equally well. It is an objective measure in that we know how it is calculated for every single student. We can audit the scores

With the "personality score" we have proof that the score was intentionally created to lower the chances of Asians getting through. Moreover there is no way to audit the scores of a particular student. They do not explain how a score is calculated. Moreover, Harvard alumni scored Asians on par with their other peers, meaning that the metric was created almost certainly to discriminate

What if that number happens to be zero?

Reply on news.ycombinator.com