holy missing the point. their labels may be arbitrary but their biology and culture is not. To pretend like an American from European ancestry is the same as one with Hispanic ancestry is ridiculous, especially since the vast majority of Hispanics in this country come from Mexico and South America. The point is that two distinct groups are being intentionally represented together, blurring the lines between identifiable distinctions such as crime rates, which have HUGE implications on how it affects the country and political discourse
Almost no Hispanics are 100% white. Many Latino people are. At least get the groups correct if you're going to act like you know something. Central and South American people are overwhelmingly non-European (white is a Euphemism for ethnic European).
Your argument is really "you're only white if you have 0% of any other race"? That seems irrational and incoherent. If you're half white, you're white.
That is absolutely my argument. If you are half one thing and half something else, you are something else, something new. Just like in viral recombination. Just like horses and zebras making a zorse. Europeans are a distinct thing. As are Africans. Mixing them together doesn't create either of those two things. You would never say a Great Dane mixed with a Chihuahua makes a Chihuahua or a Great Dane.
Which is why I made it a point to say the vast majority of Hispanics in the US are Hispanics from South America and Mexico which are not 100% European. They are distinct ethnic groups with distinct cultures and demographic characteristics. You want to merge them together because the fact that race is real and genetic/cultural differences (like crime rates) exist gives you cognitive dissonance.
Except Italian and Irish immigrants evolved from Europe whereas Mexicans and South Americans originate from the Americas. But, Anglo Saxons, Irish and Italians do have different ethnic makeups.
How hard would it be to run your program and separate Middle Eastern and South Asians from the White classification? Also was there any corrections with those misclassified and the crimes they were being charged with?
1. Very difficult or impossible. This simply isn't measured in the DOC databases and the proxies (middle eastern surnames, independent race classifiers) are insufficiently accurate on their own to be a guarantee.
2. I have this data, but didn't run this analysis. The paper was already quite lengthy as is and I wanted to focus on misclassification itself. The listed crime data was also messy. There was often more than one crime per criminal, crimes can be split into many various subtypes, etc. I had to use LLMs to manually comb though all this and label it, the result was noisy.
I wonder if this is might not be primarily because fully-assimilated Hispanics (assimilation demonstrated primarily by unaccented American English) are generally viewed by American society as being basically White. It would be very interesting to have data about the speech patterns of all of the people in the database, and see how those correlate with the race assignments.
Black people, being phenotypically more distinct, might suffer misclassification based on speech patterns less, but it wouldn't surprise me if there is still some effect. While Black American speech patterns are diverse, there are distinctly Black accents.
It's almost like the concept of whiteness and hispanic are arbitrary and incoherent. You really did all this work and managed to avoid the obvious?
holy missing the point. their labels may be arbitrary but their biology and culture is not. To pretend like an American from European ancestry is the same as one with Hispanic ancestry is ridiculous, especially since the vast majority of Hispanics in this country come from Mexico and South America. The point is that two distinct groups are being intentionally represented together, blurring the lines between identifiable distinctions such as crime rates, which have HUGE implications on how it affects the country and political discourse
They aren't distinct groups. Very large amounts of "hispanic" people are 100% or nearly 100% european.
Refer to my previous comment.
Almost no Hispanics are 100% white. Many Latino people are. At least get the groups correct if you're going to act like you know something. Central and South American people are overwhelmingly non-European (white is a Euphemism for ethnic European).
Your argument is really "you're only white if you have 0% of any other race"? That seems irrational and incoherent. If you're half white, you're white.
That is absolutely my argument. If you are half one thing and half something else, you are something else, something new. Just like in viral recombination. Just like horses and zebras making a zorse. Europeans are a distinct thing. As are Africans. Mixing them together doesn't create either of those two things. You would never say a Great Dane mixed with a Chihuahua makes a Chihuahua or a Great Dane.
Which is why I made it a point to say the vast majority of Hispanics in the US are Hispanics from South America and Mexico which are not 100% European. They are distinct ethnic groups with distinct cultures and demographic characteristics. You want to merge them together because the fact that race is real and genetic/cultural differences (like crime rates) exist gives you cognitive dissonance.
You sound like the kind of person that would have said Italian and Irish immigrants weren't white, using the same arguments. Think on that.
Except Italian and Irish immigrants evolved from Europe whereas Mexicans and South Americans originate from the Americas. But, Anglo Saxons, Irish and Italians do have different ethnic makeups.
South Americans are almost all Spanish in descent. That's white.
Really fantastic work. Thank you for doing it!
How hard would it be to run your program and separate Middle Eastern and South Asians from the White classification? Also was there any corrections with those misclassified and the crimes they were being charged with?
1. Very difficult or impossible. This simply isn't measured in the DOC databases and the proxies (middle eastern surnames, independent race classifiers) are insufficiently accurate on their own to be a guarantee.
2. I have this data, but didn't run this analysis. The paper was already quite lengthy as is and I wanted to focus on misclassification itself. The listed crime data was also messy. There was often more than one crime per criminal, crimes can be split into many various subtypes, etc. I had to use LLMs to manually comb though all this and label it, the result was noisy.
Why leave out Asians/East Indians from this? Because it makes Whites look bad in comparison?
Nebraska’s Gregory jones confirms what I always suspected: turning a white guy upside down makes him 99% likely to be a black guy
I wonder if this is might not be primarily because fully-assimilated Hispanics (assimilation demonstrated primarily by unaccented American English) are generally viewed by American society as being basically White. It would be very interesting to have data about the speech patterns of all of the people in the database, and see how those correlate with the race assignments.
Black people, being phenotypically more distinct, might suffer misclassification based on speech patterns less, but it wouldn't surprise me if there is still some effect. While Black American speech patterns are diverse, there are distinctly Black accents.
It’s all arbitrary and fake. There’s no fixing of a social construct. Remove it