I agree that that is an interesting question. The claim is that the heritability of Blackness, plus the systemic racism, is a better hypothesis than the heritability of genetically determined criminality. So the model has to have a population in which Blackness is heritable instead of having two noninteracting subpopulations of Blacks and Whites.
Or we could do it the other way and just look at the real world heritability of criminality among Blacks on the assumption that all Blacks suffer equally from systemic racism.
“... heritability of genetically determined criminality.”
I fail to see where that conclusion comes from, except perhaps implied in the first exchange. Rather the correlation of race and criminality I’ve always understood is due to the heritability of poor impulse control and other behaviors which translate to increased violent social interactions defined in this society as criminal in nature. Difference is subtle, I admit, but allows an understanding of how such inherited behavior may be much more normal/accepted in other societies and therefore genetically selected for.
I lost sight of the main issue and drifted away from it a bit. So far as I can tell, the main issue is: when we observe that a Black person in the US is more likely to be arrested than a White person, should be blame the people being arrested or blame the police arresting them? The systemic racism story blames the police. Claims about the heritability of criminality blames the Blacks, if that heritability is genetic.
So studying heritability of criminality restricting to a Black subpopulation doesn't help settle the issue, and studying heritability of criminality among the almost entirely White Swedes doesn't help either.
Criminality is heritable among the Swedes where systemic racism can't drive it because they are all the same race.
Criminality is heritable among Blacks where systemic racism can't drive it because by construction they are all the same race.
Criminality is heritable to about the same extent in a population with a mixture of races.
In this situation we don't need systemic racism to explain the last observation. Studying the racially homogenous groups is relevant to reaching that conclusion.
Apr 27, 2023·edited Apr 27, 2023Liked by Emil O. W. Kirkegaard
Repeated sentence in the last paragraph. "Note that randomness in the justice system deflates the heritability estimates (true value 50%, observed at 0% bias is about 40%). Note that randomness in the justice system deflates the heritability estimates." Also, "We can also show that adding a chance element to this doesn't effect matters." Should be affect. Feel free to delete this, obviously.
Yeah, not an ad hominem argument, and insults are not a good look either, but it certainly cures the boredom. I read insults with pleasure and guilt on my conscience. I am thinking maybe the appropriate section of each post should be given an "Insults" header, to appropriately partition it from the relevant arguments, relieve the guilt for readers and provide kinduva trigger warning.
Heritability studies, being family studies, of course examine the causes of variations within groups only--within Whites only or within Blacks only. Somehow you can be a PhD student at Harvard and still know nothing more than a typical normie about heritability studies. The topic of statistics is expansive, and maybe you get only a brief glimpse of heritability study methods until and through grad school. I would like to blame affirmative action, but speaking so confidently and so wrongly about a topic you know almost nothing about seems to follow mainly from being on the safe side of the debate. Nobody except the deplorables will hold you accountable.
I genuinely hope someone shares this for him. Ego plus education sometimes makes people think they are intelligent, when really they just passed the right classes and need to leave it at that.
Carr is tellingly inconsistent. In his Twitter example of a normal distribution, he employes height and explains the variation entirely in terms of multiple genetic loci. Yet for intelligence, differences across ethnic groups, which if plotted using geographic sampling of indigenous = pre-1500 populations, would undoubtedly show a quasi-normal distribution, cannot have a genetic basis.
When the phenotype has a social nature (so it is endogenous, as economists say, that is affected by individual choices) then there are additional difficulties, because the response at equilibrium of different groups may differ, depending on beliefs people have on the parameters and on the behavior of others. This consideration applies to criminal behavior, or discipline in school, or effort in career building.
This general consideration produces another argument to be considered. Economists analyze this issue under the heading of statistical discrimination. (I consider these issues in my next post in Aporia Magazine).
Sad, but your conclusion wrt Kareem Carr was what I came to simply from the reading of the first Twitter exchange. This is an old pattern often seen with mediocrities—especially in the academic field. Self promotion and over confidence seems another “inheritable” that of the race.
What would the heritability be for the whole population, Whites and Blacks?
I agree that that is an interesting question. The claim is that the heritability of Blackness, plus the systemic racism, is a better hypothesis than the heritability of genetically determined criminality. So the model has to have a population in which Blackness is heritable instead of having two noninteracting subpopulations of Blacks and Whites.
Or we could do it the other way and just look at the real world heritability of criminality among Blacks on the assumption that all Blacks suffer equally from systemic racism.
Or we could just study heritability of criminality in places like Sweden. https://pubmed.ncbi.nlm.nih.gov/25936380/ Hmm. Does that clarify things?
These are interesting issues. I should learn more stats.
“... heritability of genetically determined criminality.”
I fail to see where that conclusion comes from, except perhaps implied in the first exchange. Rather the correlation of race and criminality I’ve always understood is due to the heritability of poor impulse control and other behaviors which translate to increased violent social interactions defined in this society as criminal in nature. Difference is subtle, I admit, but allows an understanding of how such inherited behavior may be much more normal/accepted in other societies and therefore genetically selected for.
I lost sight of the main issue and drifted away from it a bit. So far as I can tell, the main issue is: when we observe that a Black person in the US is more likely to be arrested than a White person, should be blame the people being arrested or blame the police arresting them? The systemic racism story blames the police. Claims about the heritability of criminality blames the Blacks, if that heritability is genetic.
So studying heritability of criminality restricting to a Black subpopulation doesn't help settle the issue, and studying heritability of criminality among the almost entirely White Swedes doesn't help either.
Darn, I disagree with what I wrote.
Suppose, hypothetically, that we observe:
Criminality is heritable among the Swedes where systemic racism can't drive it because they are all the same race.
Criminality is heritable among Blacks where systemic racism can't drive it because by construction they are all the same race.
Criminality is heritable to about the same extent in a population with a mixture of races.
In this situation we don't need systemic racism to explain the last observation. Studying the racially homogenous groups is relevant to reaching that conclusion.
Or we could look at the genes directly and skip the intermediate steps. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4306065/
Repeated sentence in the last paragraph. "Note that randomness in the justice system deflates the heritability estimates (true value 50%, observed at 0% bias is about 40%). Note that randomness in the justice system deflates the heritability estimates." Also, "We can also show that adding a chance element to this doesn't effect matters." Should be affect. Feel free to delete this, obviously.
Fixed
The ad hominem attack is not a good look, Emil. If you think he's wrong, say so. That's enough.
That’s not an ad hominem attack, it’s just (justified imo) insults: https://laurencetennant.com/bonds/adhominem.html
Yeah, not an ad hominem argument, and insults are not a good look either, but it certainly cures the boredom. I read insults with pleasure and guilt on my conscience. I am thinking maybe the appropriate section of each post should be given an "Insults" header, to appropriately partition it from the relevant arguments, relieve the guilt for readers and provide kinduva trigger warning.
A verbal smack in the (smug) face seems appropriate. Hell, the Germans even have a word for it, Backpfeifengesicht.
Heritability studies, being family studies, of course examine the causes of variations within groups only--within Whites only or within Blacks only. Somehow you can be a PhD student at Harvard and still know nothing more than a typical normie about heritability studies. The topic of statistics is expansive, and maybe you get only a brief glimpse of heritability study methods until and through grad school. I would like to blame affirmative action, but speaking so confidently and so wrongly about a topic you know almost nothing about seems to follow mainly from being on the safe side of the debate. Nobody except the deplorables will hold you accountable.
I just found that Kareem Carr is "a PhD candidate in Biostatistics at Harvard," not just statistics. Well, even biostatistics is expansive.
I genuinely hope someone shares this for him. Ego plus education sometimes makes people think they are intelligent, when really they just passed the right classes and need to leave it at that.
Carr is tellingly inconsistent. In his Twitter example of a normal distribution, he employes height and explains the variation entirely in terms of multiple genetic loci. Yet for intelligence, differences across ethnic groups, which if plotted using geographic sampling of indigenous = pre-1500 populations, would undoubtedly show a quasi-normal distribution, cannot have a genetic basis.
https://twitter.com/kareem_carr/status/1651999301571162142
When the phenotype has a social nature (so it is endogenous, as economists say, that is affected by individual choices) then there are additional difficulties, because the response at equilibrium of different groups may differ, depending on beliefs people have on the parameters and on the behavior of others. This consideration applies to criminal behavior, or discipline in school, or effort in career building.
This general consideration produces another argument to be considered. Economists analyze this issue under the heading of statistical discrimination. (I consider these issues in my next post in Aporia Magazine).
Sad, but your conclusion wrt Kareem Carr was what I came to simply from the reading of the first Twitter exchange. This is an old pattern often seen with mediocrities—especially in the academic field. Self promotion and over confidence seems another “inheritable” that of the race.