Sociologists interested in the effects of genes on complex social outcomes claim environmental conditions structure when and how genes matter, but they have only studied environmental moderation of genetic effects on single traits at a time (gene-by-environment interactions). In this article, we propose that the social environment can also transform the genetic link between two traits. Taking the relationship between educational attainment and smoking as an exemplary case, we use genome-wide methods to examine whether genetic variants linked to education are also linked to smoking, and whether the strength of this relationship varies across birth cohorts. Results suggest that the genetic relationship between education and smoking is stronger among U.S. adults born between 1974 and 1983 than among those born between 1920 and 1959. These results are supported by replication in additional data from the United Kingdom. Environmental conditions that differ across birth cohorts may result in the bundling of genetic effects on multiple outcomes, as anticipated by classic cohort theory. We introduce genetic correlation-by-environment interaction [(rG)xE] as a sociologically-informed model that will become especially useful as data for more well-powered analyses become available.

Women’s opportunities have been profoundly altered over the past century by reductions in the social and structural constraints that limit women’s educational attainment. Do social constraints manifest as a suppressing influence on genetic indicators of potential, and if so, did equalizing opportunity mean equalizing the role of genetics? We address this with three cohort studies: the Wisconsin Longitudinal Study (WLS; birth years 1939 to 1940), the Health and Retirement Study, and the National Longitudinal Study of Adolescent Health (Add Health; birth years 1975 to 1982). These studies include a “polygenic score” for educational attainment, providing a novel opportunity to explore this question. We find that within the WLS cohort, the relationship between genetics and educational outcomes is weaker for women than for men. However, as opportunities changed in the 1970s and 1980s, and many middle-aged women went back to school, the relationship between genetic factors and education strengthened for women as they aged. Furthermore, utilizing the HRS and Add Health, we find that as constraints limiting women’s educational attainment declined, gender differences in the relationship between genetics and educational outcomes weakened. We demonstrate that genetic influence must be understood through the lens of historical change, the life course, and social structures like gender.