Parents of Daughters More Likely to Identify with Republican Party, But Only If High-Status

A newly published study by sociologists Dalton Conley (New York University) and Emily Rauscher (University of Kansas) examines whether biological daughters aﬀect political party identiﬁcation, traditional views of women, or opinions about abortion and teen sex.

Previous research found that having daughters promotes progressive views regarding women's issues and makes parents more likely to identify with left-leaning political parties. It is hypothesized that daughters make their parents more sensitive to women's issues. Indeed, socialization, and contact theory more specifically, propose that being exposed to individuals of a given group tends to make us more sympathetic to the culture, tendencies, and needs of this group (assuming that the interaction takes place in a non-hierarchical context). But these studies suffered from methodological problems (selection bias) and raised theoretical concerns as well by lumping together biological and non-biological children and by failing to recognize that parents may invest diﬀerently in male and female children depending on their social status (this is known as the Trivers-Willard hypothesis).

To address these concerns, Conley and Rauscher used data from 661 respondents with biological children interviewed for the 1994 General Social Survey. The 1994 survey is the only one that includes questions about the sex and birth order of a respondent’s biological children. Controlling for gender, religion, age of eldest biological child, education, marital status, size of city of residence, and generation age gap, Conley and Rauscher found that the higher the proportion of biological daughters, the more likely parents were to identify with the Republican Party: "Compared to those with no daughters, parents with all daughters are 14% less likely to identify as a Democrat ... Parents who had a daughter first are 6% less likely to identify as a Democrat. Similarly, parents who had a daughter first are 6% more likely to identify as Republican and parents with all daughters are 11% more likely to identify as a Republican than parents with no daughters." However, this effect held true only for high-status parents.

Conley and Rauscher also found that female children is negatively associated with traditional views of women, but the effect failed to reach statistical significance. Similarly "daughters are related to more pro-choice abortion attitudes and disapproval of teen sex, but these results are less consistent than those for party identiﬁcation."