Discrimination against female candidates in 2016 congressional elections: the role of modern sexism in vote choice

This project will examine the effects of modern sexist attitudes on voting behavior in the 2016 House and Senate elections in which a man was running against a woman. This project contributes to existing literature on sexism, effects of candidate gender, and voting behavior, by looking at the role of sexism in lower federal elections in the context of having a female two party presidential candidate running in the general election. Using the 2016 American National Election Study (ANES) data, variables were coded to makeup a modern sexism battery, and controls were implemented to account for the female candidate’s party, her incumbency status, and the office she was running for. Additionally, control variables were added to account for voter age, party ID, gender (male), race (white), education level, income bracket, how many children they had living with them under 18, if they were married, and their religiosity. A modern sexism scale was created by way of an additive index of sexism based questions in the ANES dataset. First, an OLS regression analysis was run to determine which voter characteristic variables were related to higher levels of sexism. Next, regressions were run to examine the influence of sexism on voting for the female candidate—my main hypothesis—which there was supporting evidence for. For House elections, every point increase on the sexism scale is associated with a .9% decrease of probability of voting for the woman candidate. When it comes to voting for a female Senate candidate, there is a .7% probability decrease with every point increase on the sexism scale