Now, a score on a single exam is not a particularly great way to admit people for many reasons, but if we accept that it is an objective test of mathematical and verbal reasoning skills, why not help minorities score high enough (or let people know that there is an exam in the first place)?

In college admissions (at least in theory), affirmative action takes effect only after all the candidates have passed some minimum bar in terms of quantifiable things like SATs, grades, etc. Admission to Stuy is based only on a test so one can apply the same principle: Look at people holistically only after they've passed the minimum bar.

I think we should help minorities make the cutoff and then take a closer look at those applicants who scored high enough (though perhaps not using a hard threshold). Of course this is easier said than done, but I think ignoring such a test completely and relying on GPAs or interviews or other things alone would not be a good idea. Most people haven't "done" many things in middle school the way people do things in high school that distinguish them when they apply to college. Please correct me if I'm wrong on this. (And I actually think one of the goals is to make sure that kids have a lot more opportunities and encouragement to distinguish themselves by the time they finish middle school)

Now, perhaps the current cutoff for getting into Stuy is analogous to getting above a 2300 on the SAT, and elite institutions seem to agree that a 2300 is no more predictive of success than a 2200 or a 2100. So it would be interesting to measure how scores on the admission exam correlate to various metrics of success in college and later in life in order to determine a similar threshold above which the score doesn't matter anymore.

   I attended Stuyvesant in 1973-76, which I learn from this article was
   the era of the highest black enrollment. I have a few quibbles: the
   article states that admission was solely by test score. In fact, one of
   my best (black) friends at Stuyvesant sheepishly told me that he
   scored somewhat below the cutoff but was admitted as part of a
   program to admit students from “disadvantaged” backgrounds with lower
   scores. He was sheepish because he knew that I knew that he was in
   fact quite wealthy and was from a privileged background. His
   father was a famous radio personality. But the system treated him as
   a member of a “disadvantaged” group because of his complexion.

   The article says “Students take the exam in October of their eighth-grade
   year”. Many students who will wind up at Stuyvesant never go
   to eighth grade. Like me, they are placed in an accelerated
   junior high program that consists of seventh and ninth grades only.
   Also, the article refers repeatedly to “middle school”. The first
   time I heard this term was after I moved away from New York.

   The article refers to “free preparatory programs”. Never heard of
   them, nor any other type of test preparation or practice. The test
   resembles an SAT. If you know English and your junior high math, it’s
   not that difficult to get in to Stuyvesant, frankly. I know this
   because I was not very good at taking tests, and was somewhat
   surprised that I got in.


   A bigger obstacle is the desire to go there. Most American students
   don’t want to face a year each of physics, biology, and chemistry,
   math up to (and optionally including) calculus, and much more,
   including economics, metal shop, and drafting. It might not sound
   like much to the Germans and Japanese, but it does not resemble the
   typical American highschool curriculum. I was shocked to meet people
   in college whose mathematical knowledge seemed to stop at what I had
   learned in 10th grade and who had never heard of mitochondria.  All
   this, and the nerd stigma, was what kept my siblings, for example,
   from taking the test and settling for our abysmal neighborhood
   school. Note that most of Stuyvesant’s population consists of the
   offspring of immigrants, who seem to appreciate the importance of
   education.

Start early (earlier than high school -- they went down to grades 6-8 to get qualified applicants), devote extra resources to identifying and mentoring early on (when it can still do some good), etc. It's a long pipeline, and you don't want it to stall early on.

Especially, the line where she states that shes heard enough about blacks having an easier time getting into schools.

I'm Asian and as someone who's seen affirmative action effect my life in the past 10 years. I can say that its fair for them to say that because that's the reality.

America has been obsessed about being diverse. Diversity for the sake of diversity is a failure in the country. Diversity born of necessity and natural selection is what we should be seeking out.

I say worry less about making sure we have blacks, yellows, white and bullshht. Worry more about helping anyone who's willing to learn at an early age. The single greatest determinant in whether we're successful is how curious and willing we are to learn. Our focus has been skewed toward things that just aren't as important to success.

*obviously, Im biased as these are my opinions but I've always been results orientated and less focused on ethnicity. I've traveled around a lot as rdl has as well and I've noticed that most countries don't put an emphasis on race as much as Americans do.

I'm with you on being against "lowering the bar" style affirmative action -- it's unfair to everyone, including the high achieving members of the protected group. The problem is the cultural and family structures don't exist to help intelligent and otherwise-likely-to-succeed people from certain groups progress through the very long pipeline to being successful startup founders. I'm just saying trying to apply late-stage correction (pushing for proportional representation in startup incubators, colleges, etc.) is the wrong place to work -- you can get by with much less expensive, much more fair interactions by starting a lot earlier. KIPP will probably do more for minority representation in the startup world than anything at the incubator level, but it will take 15 years.

One of the things the military has done pretty well is racial (and geographic) integration, through treating everyone basically the same (after integration), and aggressively getting rid of organizational or structural racism. That's different from startups, but does seem to support the idea that it's possible, given enough time.

Actually, most of the countries I've visited care more about ethnicity/race than the US, especially more so than silicon valley. Possibly selection bias.

Malaysia is, in my experience, one of the most racist and weirdly political-about-race countries I've ever visited. Thailand and other areas with an ethnic Chinese population often have really weird race based politics. I'm also not really sure how you count tribalism in places like Iraq/Afghanistan/Central Asia. Then there are places in the middle east where an Indian family who has lived there for 150 years is still not "local".

Race and socioeconomic class are maybe more tied in the US than in the other places (where it's more political), but it would be just as valid to say rural poor white kids from West Virginia have little exposure to quality math/science education early on, too.

Easier said then done when you consider that "start early" means being aware of opportunities (even people seeking out such opportunities may not know where to look), and "devote extra resources" may mean spending money that families don't have.