How (and Why) Sociologists Should Care About Heritability: Evidence from Misclassified Twins

We argue that despite many sociologists’ aversion to them, heritability estimates have critical policy relevance for a variety of social outcomes ranging from education to health to stratification. However, estimates have traditionally been plagued by genetic-environmental covariance, which is likely to be non-trivial and confound estimates of narrow-sense (additive) heritability for social and behavioral outcomes. Until recently, there has not been an effective way to address this concern and as a result, sociologists have largely dismissed the entire enterprise as methodologically flawed and ideologically-driven. Indeed, in a classic paper, Goldberger (1979) shows that by varying assumptions of the GE-covariance, a researcher can drive the estimated heritability of an outcome, such as IQ, down to zero or up close to one. Survey questions that attempt to measure directly the extent to which more genetically similar kin (such as monozygotic twins) also share more similar environmental conditions than, say, dizygotic twins, represent poor attempts to gauge a very complex underlying phenomenon of GE-covariance. Methods that rely on concordance between interviewer classification and selfreport offer similar concerns about validity. In the present study, we take advantage of a natural experiment to address this issue from another angle: Misclassification of twin zygosity in the National Longitudinal Survey of Adolescent Health (Add Health). Since such twins were reared under one ―environmental regime of similarity‖ while genetically belonging to another group, this reverses the typical GE-covariance and allows us bounded estimates of heritability for a range of outcomes of interest to medical and behavioral scientists. The Sociology of Heritability