A theory of gender’s effect on vote shift with a test based on Turkish elections

ABSTRACT

Despite great advances in the last decades, there is an enduring gender gap in a majority of the countries. Gender is prevalent in all aspects of life, and it is impossible to avoid gender roles. The choices regarding gender roles are threefold: Accept them, reject them, or have an ambivalent stand. When voters grow incongruous with their parties, they look for an alternative which is ideologically closer to them. However, evaluating this vote shift without the full effect of gender will provide an incomplete picture of electoral politics. I theorize that when voters are faced with female candidates or female party leaders, their assessment of gender roles will be a deciding factor for whether the intended vote shift will be realized. I apply this extended theory of ideological and gender role congruity to a high gender gap context, Turkey, utilizing district level electoral data. Taking into account the spatial dependence of factors that affect voting behaviour, I empirically demonstrate that for two ideologically opposing parties, the salience of gender politics through the ranking of female candidates on the closed-lists and presence of national female leadership affected their vote share in line with my theoretical expectations in the 2018 general election.

KEYWORDS:

• Ideological incongruity
• gender role congruity
• sexism
• voting behaviour
• spatial error model
• Turkey

Acknowledgements

Earlier versions of this paper were presented at the 2022 Southern Political Science Association Meeting and the 2022 Texas Comparative Politics Circle Meeting. I would like to thank Michelle Taylor-Robinson for her mentorship and Sabri Ciftci for his guidance. I would also like to thank the editor of the journal and the three anonymous reviewers for their helpful comments. The code and data needed to replicate the quantitative analyses are deposited at Harvard Dataverse. DOI: https://doi.org/10.7910/DVN/FSAFCK.

Disclosure statement

No potential conflict of interest was reported by the author.

Supplemental Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/13629395.2023.2194154.

Notes

1. These may include but are not limited to policy-related concerns, economic performance, candidate placement, or discontent with the anti-democratic behaviour of the party.

2. Since 1983, there is a 10 per cent national threshold in Turkey, which is one of the highest in the world.

3. The vote shares for the parliamentary parties in the 2015 and 2018 elections are: AKP 49.5 & 42.56, CHP 25.3 & 22.65, MHP 11.9 & 11.10, HDP 10.76 & 11.7, and IYI - & 9.96.

4. ‘Kadnla Erkek Eşit Olamaz’, Vatan, 20 July 2010, haber/kadinla-erkek-esit-olamaz/318006/9/Siyaset, reported in Kandiyoti (2010).

5. https://tinyurl.com/Bahceli-Aksener.

6. https://tinyurl.com/Erdogan-Aksener.

7. There is discussion about whether this increase translates into substantive representation. Looking at private bill sponsorships, Bektas and Issever-Ekinci (2019) find that right- and left-wing parliamentarians concentrate on issue areas that are associated with their ideologies and that this means an increase in substantive representation. However, Sumbas and Dinçer (2022) look at the record of AKP women parliamentarians on the issues of maternity leave and part-time work and find that emphasis on these perpetuate consensus on traditional gender roles and hence increased number of AKP women parliamentarians does not mean substantive representation for women.

8. In this sense, educational gender gap is like the infant mortality rate variable used in developing country studies. It is rightly assumed that this objective simple metric is the representation of many underlying factors that would otherwise mostly be missing in any analysis.

9. The year 1985 is the first available year when district-level educational data was collected and reported.

10. These regions are based on the NUTS-2 classification. The graph is available as supplementary material.

11. The correlation coefficient between mosque density and educational gender gap at the district level is 0.37.

12. https://hdp.org.tr/en/hdp-women-s-assemblies/8830/, accessed 20 June 2022.