Why Women and Men Vote Differently: The Gender Gap on the Sociocultural Cleavage

Working paper with Jonne Kamphorst, Liesbet Hooghe, and Gary Marks

Download the manuscript here

This paper investigates how and why the contemporary gender gap in voting emerges—specifically, why women are more likely to support GAL (Green–Alternative–Liberal) parties and less likely to support TAN (Traditional–Authoritarian–Nationalist) parties than men. Grounded in social role theory, we argue that the gender gap is not solely rooted in early-life traits or stable psychological differences between men and women. Instead, it develops through a sequence of gendered life-course experiences, especially in education and occupation, as societies transition toward knowledge-based economies. As women and men increasingly sort into different educational fields and workplace environments—fields that differ sharply in their “cultural-communicative” versus “technical-economic” content—their political outlooks evolve in systematically different ways. We conceptualize these education- and occupation-based experiences as key sites of adult political socialization, rather than mere reflections of preexisting traits.

To test this theory, we develop a stepwise empirical strategy that integrates decomposition analysis, univariate mediation, sequential mediation, causal panel estimators, and qualitative interviews. Decomposition shows that most of the gender gap arises from differences in life experiences (mediation) rather than men and women responding differently to the same experience (moderation). Mediation analyses using European Social Survey data reveal that field of study and occupation are the two strongest mediators of the gender gap—far stronger than standard variables like education level, income, or family roles. Sequential mediation confirms that these effects persist even after accounting for personality traits such as universalism. Finally, panel analyses using FEct estimators demonstrate that entering high cultural-communicative fields and working in high cultural-communicative occupations causally increase individuals' likelihood of supporting Green parties. In-depth interviews reinforce these findings by showing that people themselves view their educational and occupational environments as powerful influences on their worldview. Together, the results offer strong evidence that field of study and occupation are major drivers of the modern gender gap in voting, acting as pathways through which gendered divisions of labor shape political behavior over adulthood.

Figure 1: Numbers in parentheses indicate proportions in the sample, numbers on arrows without +/− signs represent the sub-sample averages of the outcome, the tendency of leaning green, and numbers on arrows with +/− signs indicate causal effect (conditional ATT) estimates generated by FEct. The numbers in brackets present the 95% confidence intervals for these estimates.