Mailbag: Flawed Race Troll Rhethoric vs. Confounders

Time to respond to yet another reader feedback email! If you want to send me a question, comment or any other kind of feedback, please do so using the contact form on the about page. For more answers to feedback emails, see the mailbag category.

Most race realists I encounter online hesitate to directly state their opinions of African-Americans (or other ethnic minorities). If they did, it would be pretty obvious that their position is based on nothing more than lazy stereotyping and ingrained prejudice. This means that they have to attempt to dress up their delusional beliefs in scientific terms to make it appear respectable. As a result, they tend to spread misinformation about scientific topics such as heritability, genetic risk factors, intelligence and aggression. Typically, they also add condescending complains about “cultural Marxism” or “suffocating political correctness gone mad” or something similar. Debunking these allegedly “sophisticated” race realists often demand both deep and broad knowledge covering topics such as adoption and twin studies, genome-wide association studies, haplogroups, principal component analysis, heritability, statistics and the biological influences of behavior. In addition, a lot of patience.

Ever so rarely, I come across what we might call an unsophisticated race realist. This simplified archetype typically do not bother to attempt to discuss the details of human genetic diversity or attempt any of the traditional pseudoscientific gambits. Instead, this kind of race realist just blurts out their favorite stereotype about a given ethnic minority. Recently, I got a feedback message from one such race realist, going by the name of Neil Gordon. After reading my detail scientific take-down of race realist assertions in the comment section of a previous post in the mailbag series, he sent me the following feedback message:

When you argued against race realism in your 2012-03-04 post, and when you debated “Rich” in the comments, you went into fine detail exploring certain questions:

* How well/badly traditional racial categories map onto the existing pattern of human genetic variation.

* Whether “Lewontin’s fallacy” is really a fallacy.

However, your decision to focus on this aspect looks to me like a delaying tactic to avoid getting to the heart of the matter.

In other words, Gordon does not seem to appreciate careful refutations of race realist misunderstandings of the scientific literature. Instead, he refers to this as a “delaying tactic”. In reality, the failure of traditional racial categories to map to human genetic variation and the fact that most variation exists within groups rather than between them is of utmost relevance to the validity of race realism. By annihilating the foundation of race realism — their misunderstandings of scientific research — the game is over. Race realism collapses like a house of cards and nothing more remains besides the prejudice and lazy stereotyping.

Anyhow, what is this “heart of the matter” that Gordon alludes to? As it turns out, it is nothing more than prejudice and lazy stereotyping. What a surprise…not.

The core of race realism – the foundation for the policy proposals – is essentially this:

P: Certain groups of people, e.g. Americans with mostly African ancestry, have lower intelligence and a greater propensity to violent crime and corruption than other groups e.g. Americans with mostly European ancestry. Moreover, this is explained to a significant extent by the genetic differences between the groups.

Gordon is essentially advancing the mistaken position that African-Americans are somehow less intelligent and more violent than European-Americans and that genetic differences is the most important explanation for these alleged “fact”. What evidence does Gordon present for these claims? Predictably, he does not present any evidence at all for his assertions. No references to the scientific literature. Not even a link to a news article about published research.

As we shall see, these claims do not hold up once we introduce evidence from published studies. It turns out that the alleged IQ gap between African-Americans and European-Americans has been closing due improved education and environment and once you take things like socioeconomic status and other factors into consideration, the difference in average IQ or violent crime is no that large. Or to summarize:

Ever heard of confounders?

For an easy introduction to confounders, consider the risk of Down’s syndrome and birth order. Correlation studies have been done and it seems that a higher birth order is associated with an increase in risk of Down’s syndrome. So the first child has the lowest risk, the second higher, the third higher than that and so on. Does this mean that birth order causes Down’s syndrome? As it turns out, no. This is because there is a confounder lurking in the shadows: the real relationship is between maternal age and the risk of Down’s syndrome. It just so happens that children born to older mothers tend to have a high birth order. To alleged causal relationship between birth order and the risk of Down’s syndrome is illusory.

A similar issue exists for ethnic group status and intelligence. The correlation between intelligence and socio-economic status of the family is substantial, around r = 0.4, the IQ gap was narrowed over time and adoption studies and early interventions have shown that IQ can be improved by as much as 15-20 points. Such early intervention studies have also shown that individuals are considerably less likely to be arrested multiple times, less likely to end up on welfare, more likely to graduate on time, more likely to acquire a 2000 USD+ income per month and more likely to have their own home compared with non-intervention controls (Passer et. al. 2009, p. 99, pp. 459-462).

As for ethnic group status and crime, an instructive example is data on immigrant crime in Sweden. If you do not control for confounders, the observed over-representation is around 2.5. This over-representation also needs to be put into the sociological context: people whose family ever has been on welfare have an over-representation of 6.1, people without a high-school education have an over-representation of 5.7 and people with a very low-income have an over-representation of around 5 (Martens och Holmberg, 2005). So even if you do not control for confounders, the immigration over-representation is considerably smaller than that of poor and uneducated, but obviously there are no race realist who thinks that poor and uneducated native Swedes should be deported.

Finally, in a stunning admission, Gordon states the following:

It should be obvious that whether within-group variation is greater or less than between-group variation is irrelevant to the claim above, as long as you’re not disputing that there’s some amount of between-group variation. Likewise, whether genetic variation across the whole human species is better pictured as discrete ‘clusters’ or some smooth geometric shape is irrelevant to the claim above, as long as you’re not disputing that African-Americans and European-Americans are sufficiently distinct to make the statement P above meaningful.

In other words, Gordon explicitly concedes that the debating tactics and tropes commonly used by race realists are essentially irrelevant to the core beliefs of many race realists. This is essentially a confession that race realism, as typically encountered on the Internet, is based on pseudoscience and misuses of the scientific literature.

References:

Passer, M., Smith, R., Holt, N., Bremner, A., Sutherland, E., & Vliek, M. (2009). Psychology: The Science of Mind and Behavior. New York: McGraw-Hill Education.

Martens, P. och Holmberg, S. (2005). Brottslighet bland personer födda i Sverige och i utlandet. Brottsförebyggande Rådet. Accessed: 2013-11-23.