>when you see a random black person and a white one and an asian, and a jew, your average racist will factually be right in guessing their relative intelligence
Wrong. Individuals are not populations. The most you could say is that if someone guesses a large number of times they'll be right slightly more often than not. This is not a trivial point.
I get where you're coming from, racism sucks.
Statistically the odds are different, stereotypes are usually pretty accurate.
Think recruitment and probability, let's say it cost you $200/hr to screen a candidate from CV to interview and you have $1000. We have law to ban it as an assessment criteria, on paper. It's also morally wrong (at level 3 or level 4 of the moral pyramid).
Warm fuzzy feeling is just that, politically correct is easy. Personally I think nationalism is stupidity, but that's our history and who we are. I feel comments like these is head burying in some way, that masks us from the actual fixable problems like unconscious bias. How about we acknowledge the act of giving equal opportunity to assess people individually as a good act of social responsibility while we're still trying to fix environment and education for the less fortunate.
You miss my point which has nothing to do with warm fuzzies or political correctness.
I'm not accepting as fact that there are innate intelligence differences, but for the sake of argument let's postulate small differences between race/ethnic/gender populations. When you screen that way you throw away almost half your candidate pool for no good reason. You achieve worse results than if you use some other, not quite as lousy, method of culling. You'd probably have to work at it to find as weak a correlation. Sexism and racism aren't only evil, they're stupid too - even if the tiny intelligence differences that disputed authors claim to have found were real.
So in what way are stereotypes "pretty accurate" in the context of hiring? You aren't hiring a population so small population differences aren't super useful - except perhaps at the tails of the distribution. If you attract just the cream of the crop, you might have an argument. That population is likely to be very skewed.
Wrong. Individuals are not populations. The most you could say is that if someone guesses a large number of times they'll be right slightly more often than not. This is not a trivial point.