Rufo described Jonatan Pallesen as “a Danish data scientist who has raised new questions about Claudine Gay’s use – and potential misuse – of data in her PhD thesis” in an interview published in his newsletter and on the Manhattan Institute’s City Journal website last Friday.
He did not tell readers that a paper featuring Pallesen’s own statistical work in collaboration with the eugenicist researchers has been subject to scathing expert criticism for its faulty methods, and characterized as white nationalism by another academic critic.
The revelations once again raise questions about the willingness of Rufo – a major ally of Ron DeSantis and powerful culture warrior in Republican politics – to cultivate extremists in the course of his political crusades.
The Guardian emailed Rufo to ask about his repeated platforming of extremists, and asked both Rufo and the Manhattan Institute’s communications office whether they had vetted Pallesen before publishing the interview. Neither responded.
Yes, it is highly controversial, and rightly so. First, an IQ is a number that is based on an intelligence test and intended to measure an individual’s cognitive ability in comparison with a reference population, typically with other people of similar ages and in the same country (i.e., the population that they belong to). Intelligence tests are meaningless for group comparisons such as comparisons between countries or ethnic/religious groups, and doing so represents a misuse and misinterpretation of IQ scores. Researchers are not “biased” against this based on their political opinions. They simply object to the objectionable use of these tests.
Second, group comparisons about intelligence are also problematic for a variety of other reasons, and studies that claim to find group differences tend to conflate them with other between-group differences (e.g., different socioeconomic, nutritional, educational influences, among others). These studies are essentially pseudo-science.
Finally, although genetics do seem to play a significant role for cognitive ability, it’s important to realize that statements like “IQ is x% heritable” are statistical estimates. These estimates are obtained by comparing sources of variance that can be attributed to shared vs. non-shared genetic and environmental influences. As such, any heritability estimate is specific to its social context (e.g., countries). In fact, heritability estimates tend to be higher in more equitable societies, because they reduce the impact of environmental influences (e.g., wealth, parental education), thus increasing the relative proportion of variance that can be attributed to genetics (but obviously genetics in, say, Sweden still work the same as they do in the US).