[wpietri]

As a white person in the US, I'll tell you true: my opinion on racism is much less well informed than the average non-white person. Because a) I almost never experience it, b) I am less inclined to notice it, and c) non-white people are very unlikely to talk to me about their experiences and thoughts on the topic.

So yes, a person's race is entirely relevant to considering the quality of their opinions on racism. For basically the same reason that I will happily weight a person's opinions on software development based on whether they're an experienced coder or a non-technical manager. Experience matters.

[gizmo]

I agree with the statistics (racism is real; white people are less like to be discriminated against or experience indirect racism) but draw the opposite conclusion.

When dealing with issues of discrimination the point is to treat everybody equally, to treat people based on their merits and not based on the color of their skin or the statistics of their demographic. If we conclude from research that one race is more intelligent than another, or taller than another, or has more aptitude in some domain, then a rational agent would reason: "well, this person is from domain D, and people from domain D have less aptitude in Y, and since I am hiring people for that aptitude, I am less inclined to hire this person.". Logically sound, but still discriminatory. It's discriminatory because the person is treated based on the demographic he belongs to, not on his or her individual merits. So even if the research points out that one demographic is inferior to another demographic on one axis that still doesn't justify treating an individual based on his membership of a demographic. To put it bluntly, if research shows (it doesn't) that white people are absolutely beyond a shadow of a doubt better at programming than black people discrimination against black people would still be wrong. The differences in aptitude between the individuals always trump differences in aptitude between large groups.

So when you say "a person's race is entirely relevant", I think you're completely and terribly wrong. Because it's irrelevant whether a typical white person understands less about racism than a typical black person. What matters is whether the specific individual understands. Again, the individual is not to be judged based on his membership of a demographic.

Yes, a typical white person understands less about racism than a typical black person. And a majority of the prison population consists of black people. And if you apply those generic traits of the demographic to the individual you're racist. When you judge a black person on his statistical likelihood of having a criminal record then that's racist. When you judge a white person on his statistical likelihood of ignorance on race issues that's also racist.

[mck8]

It seems likely that k33n's opinion regarding racism in the world is, like most opinions, based largely on personal observations and experiences rather than a careful consideration of all available data. He does not observe racism to be a significant problem and thus concludes that racism is not prevalent or relevant today. You argue that k33n's race is irrelevant to his reasoning.

Both of you are incorrect. We have two hypotheses: racism either is or is not a relevant problem. And, we have that k33n does not observe much racism. We (and k33n) should update our beliefs based on the ratio P(k33n does not observe racism|racism is not relevant)/P(k33n does not observe racism|racism is relevant). The value of this ratio depends on k33n's race. If k33n is white, he's unlikely to observe much racism in either case, and the ratio's value is close to 1. If k33n is black, he's much more likely to not observe racism if racism is not prevalent, so the ratio's value is high and useful for updating beliefs.

You're also arguing that we ought not conclude that someone cannot understand racism simply because they are white. I agree with you, but I don't think that's the issue here. The issue is that k33n is white and seems to glibly conclude that racism isn't a real problem. He couldn't have reached this conclusion by looking at statistics because, as you know, the statistics push you toward the opposition conclusion. So, he probably came to this conclusion because he observed that racism isn't a problem for him personally. In doing so, he was being unreasonable and somewhat callous.

[wpietri]

Taking four paragraphs to get to your conclusion is a good tactic there. Because maybe by the end people will have forgotten that I'm not actually advocating judging people by their skin color.

[gizmo]

It pains me that in 2011 I still have to explain you can't justify discrimination (of any type) with statistics.

[wpietri]

That would be an interesting and relevant point if I had actually tried to justify discrimination with statistics.