[js8]

I am leftist and I agree with the last paragraph, and I think many leftists would. And that's the point - the stereotypes are useless in policy and decision making, and studying them is unscientific. They are factoids perhaps good for cocktail parties.

I don't think that existence of correct stereotypes poses a great risk for ideology of the left, unlike global warming, which actually is a risk for completely free market ideology, because there is no way free markets can deal with externalities. It's only if you accept naturalistic fallacy (that stereotyping is natural, therefore ethical) you get these problems.

I think left has ideological problem (and thus bias), but somewhere else. Left cannot very well deal with people who perceive risk differently. Which is actually kinda connected to the existence of stereotypes.

[mr_overalls]

Let's say that you're forecasting budget requirements for a police department in a city whose demographics are changing.

Research indicates that in the United States, the correlation between violent crime rates and percentages of US state populations that are Black and Hispanic is 0.81. Controlling for poverty, education, and unemployment only reduces this to 0.78. [1]

Wouldn't this unusually high correlation be useful in making your budgeting decision?

[1] http://www.colorofcrime.com/colorofcrime2005.html

[js8]

No. Forecast based on the crime rate trend would be more precise than using race as a proxy.