The Arnaz-Ball effect?

In the last post, we saw that the Hispanic-non Hispanic White score gap was 0.8 SD in the NLSY 97.* The majority of Hispanics are Mexican and this differential is no less than that between non-Hispanic Whites in the US and Mexicans in Mexico. We also saw that the gap between third generation Hispanics and third generation non-Hispanic Whites was 0.3 standard deviations less than this. Some might interpret this generational difference in purely environmental terms. But “Hispanic” is an ethnolingual, not racial, designation. It’s inclusive except when the MSM self-servingly deems otherwise. George Zimmerman, for example, is properly Hispanic not in part but in full because his mother was Hispanic and because Hispanicity is not a linear function of blood quantum. Given the non trivial intermarriage rate between non-Hispanic Whites and Hispanics, one might propose that part of the Hispanic generational effect is genetically conditioned via admixture. Such a hypothesis would predict that the Hispanic offspring of two Hispanics would perform worse than the Hispanic offspring of mixed ethnic parentage. To investigate, I returned to the Add health study and decomposed scores by parentage. The variables are noted below. A statistically significant difference was found between third generation Hispanics who had two Hispanic parents and those who had one Hispanic and one non-Hispanic White parent. Third generation Hispanics with two Hispanic parents performed no different from second generation Hispanics. It’s notable that more than one third of 3rd generation Hispanics in this sample have a non-Hispanic White parent, while, based on marriage demographics, one would expect less than one fifth. The discrepancy might be explained by sampling bias, resulting from parental non responses.

One might point out that the offspring of one non-Hispanic White and one Hispanic parent perform little different from those of two non-Hispanic White parents. But this is accountable, from a genetic perspective, by the fact that most non Hispanic White-Hispanic parings involve Hispanics who identify, racially, as white and who, presumably, have a higher degree of Caucasoid admixture (or, possibly, benefited from a higher degree of cross assortative mating for intelligence and Caucasoid phenotypes.) Consistent with this interpretation, self identifying White Hispanics in this sample slightly out-perform self identifying non-white Hispanics.

This, admittedly, is weak evidence, given the sample size and the cultural loadedness of the assessment, for what we might term a Arnaz-Ball effect, named after the comedic couple. But it’s corroborated by statistically significant correlations between Caucasian phenotypes and indexes of IQ, which I was able to find amongst first, second, and, importantly, third generation Hispanic immigrants in both the (nationally representative) Add health and NLSY 97 samples. Below are the relevant correlations concerning third generation Hispanics. (The color scales used in the NLSY97 and ADD health run in opposite directions, so despite the differences in signs (below) the association goes in the same: lighter complexion, higher IQ.) To note, The magnitudes of these are no less than those of the respective correlations found in the African American population (refer to my pioneering, if perfunctory posts on “colorism“.)

National Longitudinal Survey of Youth 1997

The National Longitudinal Study of Adolescent Health

*“Hispanic” is an ethnolingual designation which describes those with cultural and linguistic ties to Latin America; generally, it describes people who have recent Latin American ancestry. “White,” on the other hand, is an ethnoracial or racial designation, which describes, usage depending, either individual with substantial cultural and genetic ties to historic Europe or those who are of, primarily, West Eurasian stock. The two groupings are not exclusive and overlap. Some Hispanics are White and some Whites are Hispanics. When comparing Hispanics to Whites we are making an aquatic animals to mammals comparison, which is permissible when dealing with mean differences. Sometimes it’s informative to distinguish non-Hispanic Whites from all Whites and to compare them to Hispanics. This is equivalent to comparing non-aquatic mammals to aquatic animals. Properly, the two groups under discussion are “non-Hispanics Whites” and “Hispanics,” but the modifier “non-Hispanic” is cumbersome, so sometimes “White” is used to designate “non-Hispanic White.”

About these ads