At the Intersection of Gender and Language: Why do Francophone Women have Lower Levels of Political Participation?

ABSTRACT

In Canada, research has found few differences in levels of political participation between Francophones and Anglophones despite traditional differences in resources leading to participation. This study argues that it is when gender is considered alongside language that differences emerge. Differences in conditions between Francophone and Anglophone women may mean that the explanations for the gender gaps differ. Findings show that Francophone women in Québec have lower levels of political participation than Anglophone women and men across Canada. These gender gaps are small but significant. Differences in resources and involvement in voluntary associations help account for Québec Francophone women’s lower participation. Additionally, these results indicate that the different political context and the different political influence of the women’s movement in Québec matter in determining explanations for gender gaps in political participation.

Keywords:

• Political participation
• gender gaps
• Québec politics
• women’s movement

Acknowledgments

A previous version of this paper was presented at the 2010 Congrès annuel de la Société québécoise de science politique at the Université du Québec à Montréal. The author would like to thank Geneviève Tellier, Brenda O’Neill, and the two anonymous reviewers for their helpful comments.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

^1. Official data are not available by gender.

^2. A sample weight provided by the General Social Survey is used to insure that the results reflect socio-demographically the Canadian population. The variable smpl_wght is used.

^3. Respondent’s household language is used as an indicator. Respondents who neither speak French nor English at home are excluded from the analysis.

^4. Appendix 1 contains the proper wording for each variable used in the analysis.

^5. Appendix 2 includes the percentages of males and females that have engaged in each political activity included in the index.

^6. To determine that acts of political participation included in the index are similar, Cronbach’s α is performed. The Cronbach’s α measures the internal consistency of the index. For the eight questions included in the index, the Cronbach’s α = 0.645, meaning that the scale is somewhat reliable in measuring a single variable.

^7. Since the dependent variable only has three categories, we perform ordered logistic regressions as a robustness check. Since results (not shown here) do not significantly vary from those of the OLS regressions, we only include and discuss the OLS results as they are more readily interpreted.

^8. Income levels are not included in the analysis because the variable “household income” has over 30 percent of missing cases. A missing cases analysis—not shown here—reveals a positive t-test between missing cases in income levels and the index of political participation, creating a possible bias in the analysis if income was to be included.

^9. To address this criticism of the number of hours spent at work, a double day variable is created by interacting the number of hours with the presence of children at home. It is hypothesized that citizens who spend more time at work and have children should be less likely to be involved in political activities. Results (not shown) do not differ significantly from those without the interaction term; thus, we opted to not include the interaction term.

^10. The results for work as a main activity and hours per week spent working seem to be contradictory. To verify whether the relationships in Table 3 are accurate, we check the effect of each without the other. Results (not shown) indicate that when included alone, both work as a main activity and hours spent working per week are not significant, meaning that they do not influence citizens’ levels of political participation. It is possible that the coding of working as a main activity is responsible for these results as the reference category includes part-time workers as well as students and retired citizens.

Additional information

Notes on contributors

Katrine Beauregard

Katrine Beauregard is a lecturer at the School of Politics and International Relations at the Australian National University. She completed her PhD in Political Science at the University of Calgary, Canada. Her research focuses on political participation, political institutions, and gender. Her previous work has been published in the European Journal of Political Research.