Commentary on Fuerst and Kirkegaard: Some groups have all the luck, some groups have all the pain, some groups get all the breaks

Soy extranjero en mi tierra, y no vengo a darles guerra, soy hombre trabajador. Y si no miente la historia, aqui se sento en la gloria la poderosa nacion entre guerreros valientes, indios de dos continentes, mezclados con espanol ("Somos Mas Americanos" by Los Tigres Del Norte).Fuerst and Kirkegaard have written a comprehensive paper that examines the Racial-Cognitive Ability-Socioeconomic (R-CA-S) hypothesis, a synthesis of hypotheses that broadly conceptualize racial differences in cognitive abilities as resulting from environmental, genetic/epigenetic, and evolutionary progression (e.g., Lynn 2008). This paper follows other research that examines whether cognitive abilities mediate the association between race and socioeconomic status. Collecting regional and intra-regional data is often elusive. Fuerst and Kirkegaard have not only been diligent and creative, but have also been appropriately transparent by presenting their raw data for others to replicate or reinterpret the analyses. Further, although tackling such a topic is often met with immediate knee-jerk resistance, science is built on testing a wide range of hypotheses even when they are not popular. To their credit, the authors have handled a controversial topic tactfully and respectfully.Broadly, we agree with the overall results of the findings. We concede that the reported associations among European ancestry, cognitive abilities, and socioeconomic status in the Americas are robust. European-descended groups in the United States, for instance, have both higher incomes and greater social capital with which to increase both their individual and aggregate wealth (DeNavas-Walt & Proctor, 2014) and also perform better on measures of cognitive abilities (e.g., National Center for Education Statistics, 2015). We do, however, take some exception to the interpretations proffered for their findings and the atheoretical approach the authors took toward analyzing the data. Our present commentary focuses on the conceptual problems in their paper that we believe need to be addressed to expand research on the R-CA-S hypothesis, as well as their corresponding parallels in the seemingly suboptimal statistical methodologies used. While we applaud the authors for being more transparent than most, there is a notable absence of clarity when describing some of the methods used in this paper, so our characterization of these methods as "suboptimal" might be due to their inadequate description in this paper rather than the procedures that were actually applied.Paraphrase of Main ThesisFuerst and Kirkegaard essentially argue that being genetically of European heritage provides an advantage in general cognitive abilities, and that those higher cognitive abilities lead in turn to the achievement of a higher mean level of socioeconomic achievement. We will summarize in turn our concerns regarding each of the points in this proposed causal progression.Estimation of Genetic HeritageAmong the methodological concern that we had was one regarding admixture in the Americas analyses. Fuerst and Kirkegaard obtained admixture frequencies from genetic studies that estimated the proportion of admixture present within the states of each respective country. For data points that they were unable to find, they took steps to impute estimates based on surrounding areas. For these analyses, some of the admixture estimates were estimated by dividing parental groups:In regards to hybrid identities such as Mestizo and Mulatto, percentages were split by parental group e.g., one half European and one half Amerindian. For tribrid identities such as Montubio, percentages were split three ways. Assumptions had to be made for a number of nations. (p.281).Dividing admixed populations in half (or in thirds) based on their parental groups may produce an over or underestimation of levels of racial descent. That is, the estimations derived from imputed data may be unrepresentative of the underlying distributions. …