Tuesday, July 03, 2012

Cross-national IQ, inequality, and homicide

Using data from Richard Lynn and the recent WHO study, the correlation between national level IQ and the homicide rate across 181 countries is -.51. This is a strong association, so smarter countries are much less violent. The only cross-national correlate of homicide that is this strong is income inequality (more unequal countries are more violent). In my data set, the correlation between the GINI inequality index and homicide is .48.

UPDATE: Since I mentioned inequality, any thoughts on how to explain the correlation? The inequality-homicide link is reliable in macrolevel research. Sociologists claim that a large gap between the wealthy and the poor generates anger and resentment which low-status people take out on convenient, low-risk targets. It doesn't seem plausible to me that Bill Gates is an important cause of black-on-black ghetto violence. Compared to Europe, the US has a fairly high level of inequality, but racial diversity probably explains its inequality-homicide correlation. As for the cross-national link, is it that poorly endowed populations are both violent and unlikely to create more equal arrangements? I'll have to see how much the inequality-homicide link is reduced when adjusted for mean IQ.

Back to the US question: if American liberals are so concerned about large gaps between the wealthy and the poor, why would they invite large numbers of both Chinese and Mexican immigrants? At least the immigrants of a century ago were similar (with the exception of Jews).


pat said...

This is a very interesting topic. Like most people who are actually familair with the literature I knew that IQ and violence were related. I presumed that blacks who have lower IQs and also have higher rates of violent crime had the second because of the first.

I don't remember if that arrow of causality is explicitly asserted in The Bell Curve but it was certainly implied. Low IQ led to a whole series of other problems.

Yet there are several counter examples. Trisomy 21 (Down's syndrome) and William's syndrome both lower IQ but the victims of these diseases are notably non-violent. These dull witted people are sweet and friendly.

In college there were the "A" students and the "C" students. I don't recall thinking that the "C" students were more dangerous. Controlling for race, are the "grinds" really more violent than the "brains"?

Or is there a third independent variable that causes both low IQ and high agression? Rushton seems to think so.

Pigmentation theory


pzed said...

I don't remember an explicit arrow in The Bell Curve because it's been a while, but it would be possible to check. In any longitudinal study, just look at a cohort's child/teen IQ and look at whether it was inversely related to violent crime which would presumably have occurred at a future survey.

Anonymous said...

Maybe Rindermann would be helpful to look at. From what I remember high inequality is just a function of how low the IQ of the bottom 5%. The 5th %ile in Norway is like IQ 85. People like that can function as unskilled workers and there are very few people that low in Norway. But in those countries where the bottom 5th %ile is like IQ 55, I mean, hell, even a cleaning lady is going to be miles ahead that.

Average IQ, 50th %ile is not as useful as the bottom 5th %ile and the top or 95th %ile, because places like Mexico and the USA have huge IQ differences and high inequality. And some sub Saharan places have only a medium spread yet also have high inequality. But places with very few people at the bottom are more equal because the people themselves are more equal. Hey! the identity property!

Aeoli Pera said...

"Since I mentioned inequality, any thoughts on how to explain the correlation?"

Female hypergamy? Men with children are less violent.

Just an idea.

Aeoli Pera said...

Related: Low-status men are violent because they have less chance of having sex, much less reproducing.

Aeoli Pera said...

These are conjectures, not statements. (I really need to work on my commenting etiquette.)

Anonymous said...

Related: Low-status men are violent because they have less chance of having sex, much less reproducing.

Or violent men are low status because they are violent and being incarcerated reduces their lifetime fertility.

I think Ron already posted average number of children that incarcerated men have. It was 2.3 or something close to that. It wasn't lower than other men. So, data suggests violent behavior doesn't reduce men's reproductive success. Women are not all that put off by violent guys. That is is a feminist myth. Women like masculine men, not wusses.

Vito said...

"As for the cross-national link, is it that poorly endowed populations are both violent and unlikely to create more equal arrangements?"

I think you got that backward. The "well endowed" blacks are the most violent.

Aeoli Pera said...


Your response is perfect, and I stand corrected.

One thing I was thinking about at the time (and forgot to mention) is that low-status men might become more violent in order to increase their reproductive success.

And here's why I think my idea went wrong: Men are pretty much typecast into their sociosexual stratum after puberty. At least, my experience shows there isn't a lot of mobility from class to class. I hear you can recreate your image and personality at college, but I've never seen it.

Anonymous said...

IQ and socioeconomic inequality (GINI index) are not the best predictors of national murder rates. The strongest correlate of murder at the national level is adolescent fertility. HIV rates also perform better than GINI. The explanation is mating competition theory, popular with evolutionary biologists. In a nutshell, poor socities that allow men to compete freely for women have more HIV, higher adolescent fertility, higher rape rates and higher homicide rates.

See "Cultural Differences in a Globalizing World" (Emerald, 2011) by Michael Minkov