Modeling the Cultural and Biological Inheritance of Social and Political Behavior in Twins and Nuclear Families

There has been growing interest in the possibility that biological factors, including genes, might contribute to differences in political and social behavior. Over the last 40 years, human statistical and behavioral genetics have developed a variety of models that allow the effects of genetic and non-genetic inheritance to be estimated from human kinship data. Until quite recently, these methods were unknown to political scientists and there has been no systematic attempt to illustrate the general approach to modeling genetic and social influences on differences in complex human social traits in the political science literature. The current chapter seeks to provide political scientists with the elements of a conceptual and methodological toolkit for analyzing overall biological and social influences on socially significant political outcomes. We introduce approaches that may be used to analyze family resemblance and estimate the contributions of multiple genetic and environmental sources if individual differences. The approach is illustrated by analysis of the correlations between pairs of adult spouses (N=8,287 pairs), parents and offspring (N=25,018), siblings (N=18,697), monozygotic (N=4,623) and dizygotic (N=5120) twin pairs comprising an informative subset of relationships from studies of extended kinships of twins in Virginia and Australia. The two studies comprise self-reports from a total of approximately 50,000 relatives. Six illustrative physical and behavioral variables were chosen to reflect potentially different mechanisms of familial resemblance and transmission: stature, conservative-liberal orientation; neuroticism; church attendance; political affiliation; educational attainment. The effects of assortative mating (the tendency for non-random pairing of potential mates) were incorporated in the model for biological and cultural inheritance and tests conducted for sex differences in the effects of genes and environment. The results showed that the two samples gave comparable best fitting-models for each variable. Family resemblance in stature is explained almost entirely by the additive and dominant effects of genes. The same model accounted for differences in neuroticism, though the overall genetic contribution was far smaller. By contrast, the correlations between relatives for political affiliation were entirely due to environmental influences, including a large direct influence of parental political affiliation on their offspring. The other variables showed a mixture of genetic and non-genetic influences on family resemblance. The contribution of parents to children was largely genetic, but other environmental factors shared by siblings and twins made significant contributions to variation in these outcomes. The correlation between mates for personality (neuroticism) was extremely small, modest for stature, and marked for conservatism, educational attainment, church attendance and political affiliation. All spousal correlations are positive. The final model for political affiliation is consistent with a complex of longand short-term social influences within and between families. The results for a major dimension of social attitudes (conservatism-liberalism) suggest a mixed model that implicates both social and biological influences. Data on the spouses of twins and siblings are used to test a variety of assumptions about the processes underlying the correlations between mates. The analyses strongly support the view that mate selection depends mainly on the direct choice of spouses for the measured traits (―phenotypic assortative mating‖) rather than selection for family background (―social homogamy‖) or mutual reinforcement as a result of spousal interaction. Longitudinal data are presented on the development of social attitudes in adolescent that reveal a stark contrast between the causes of individual differences among juveniles, which are entirely due to the effects of the shared environment, and those in adults in which the effects of genetic differences are much more marked. Implications and limitations of the model, design and statistical method are noted. Proposals for future study are considered.