The last 20 years (or at least the last 10 years) of focus on racial disparities has mostly missed the point, as has this article. Rather than saying, "the results are tilted, the test is discriminatory" (which made sense in 1960 and in many cases even much more recently, but fewer and fewer today), we should be asking, "why aren't they passing, and what can we do so that they DO pass?". I do not doubt that affirmative action did a tremendous amount to lead us to where we are today, but it can only take people so far. What we need to do is focus on why we are getting these results when the test is not the problem (which in this case it pretty obviously is not). Affirmative action is a cheap way to put minorities in positions but doesn't fix the underlying disparity, which the tests are showing.

Fixing that disparity requires a lot of work, but it's what the government should be focusing on rather than finding a way to ignore or get around the test scores. It's hard, politically difficult (look at how much flak Cosby gets) and expensive, but it's the real work that has yet to be done. If you think that throwing out the results every time they look bad will make minorities more equal, you're living in the past.

I am convinced that there are plenty of smart minority students out there, in fact I'd be willing to bet that some of those same minority test takers could pass that test if given the right tools or environment. The hard part is finding out what they need. Perhaps that's what they should look at these tests for - where did they have problems, and why?

I almost entirely agree with you.

I do not agree with the implication that racism is dead; it is absolutely not, and is still manifest in ugly, common ways.

Your point, however, is one I largely agree with: that a test reveals an apparently "racially" distributed performance difference in rest results does not logically mean that the test must be racially biased - unless you are a "civil rights" bigot.

Having been in the position of having to hire people, in an admittedly somewhat peculiar and obscure field, it was noteworthy to me that I never had more than 1 "black" applicant to the position (in a legitimate, sane, publically listed, medium sized corporation). He just did not have the experience. Should I have hired him? I am utterly confident that I should not.

The lack of "black" staff in the department was absolutely nothing to do with anything other than the lack of applicants, and my refusal to engage in social engineering on behalf of some abstract notion of of what the world maybe should look like.

"Fixing that disparity requires a lot of work, but it's what the government should be focusing on rather than finding a way to ignore or get around the test scores."

There is a limit to which you can "fix" disparity. Can the disparity between Ashkenazi Jews and other whites in academia be fixed? What about the disparity between whites and Chinese?

local selective pressures exerted even over several hundred years can cause big changes in cognitive function.

<link>

Go peddle your neo-Jim Crow Bell Curve nonsense elsewhere

Nor should the government "fix" disparities, if any, between Ashkenazi Jews, Chinese, and other whites. As a non-Jewish white person, I would hope the Chinese and Jewish people you reference would be rewarded according to their abilities and performance and not encumbered for the benefit of less deserving white people.