Measuring the disproportionality of Turkish elections

ABSTRACT

This article explores eight studies which calculate the disproportionality of Turkish elections held between 1961 and November 2015 using the Gallagher index. Using linear regression analysis, this article finds that authors grouping the vote and seat shares of independent candidates exaggerate the disproportionality of Turkish elections. In contrast, authors excluding the vote and seat shares of independent candidates and of minor parties and independent candidates underestimate the disproportionality of Turkish elections. Ultimately, this article concludes that grouping and excluding independent candidates, and excluding minor parties and independent candidates, are path dependent outcomes established by Gallagher (1991) and Lijphart (1994) and, to ensure the accuracy of their Gallagher indices, researchers should disaggregate the vote and seat shares of political parties and independent candidates.

KEYWORDS:

• disproportionality
• elections
• Gallagher index
• representation
• Turkey

Acknowledgement

The author would like to thank the two anonymous reviewers for their constructive feedback and Ioannis Grigoriadis and Ümit Erol Aras for their editorial assistance.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Bochsler also calculated the correlation between the accuracy of disaggregated Gallagher index analyses and Gallagher index analyses using the Lijphart index, Least-components method, logical boundaries method, and Bochsler’s own geometric model. The function for Bochsler’s (2022, 176–177) geometric model is $\mathrm{V}$_O,i = V_O,1 (1/(i −1))^g.

2. Individualized constituency-level voting data for independent candidates was sourced from the Supreme Electoral Council of Turkey for the 2007 to 2018 elections (see https://sonuc.ysk.gov.tr/sorgu and https://www.ysk.gov.tr/tr/22-temmuz-2007-xxiii-donem-milletvekili-genel-secimi/5001).

3. Disproportionality figures were sourced from Arslantaş et al. (2020, 133), Çakırözer (1995, 50, 69), Çarkoğlu and Aksen (2019, 47), Demirkol (2015, 100), Kalaycıoğlu (2002, 58), Öney and Selck (2017, 221), Tezcür (2012, 119), and Wakeman (2009, 115). Authors grouping independent candidates together include Arslantaşet al., Çarkoğlu and Aksen, Demirkol, and Tezcür, authors excluding independent candidates include Çakırözer and Wakeman, and authors excluding minor parties and independent candidates include Kalaycıoğlu and Öney and Selck. To the best of the author’s knowledge, there has been no attempt to calculate the Gallagher index of the 2018 election, and so this election was excluded from further analysis.

4. The disparity ranged from 3.19% in 1965, falling to 1.13% in 1983. At the 1965 election, this disparity partly stemmed from the use of the national remainder system which prevented independent candidates from pooling their votes at the national-level, hindering the success of independent candidates more so than when compared to political parties (Kalaycıoğlu 2002, 60).

5. The national threshold had no impact at the 1983 election given that the military only allowed three political parties to participate at the election and these parties exceeded the threshold (Çarkoğlu and Aksen 2019, 46).

6. The difference between the percentage of votes and seats received by independent candidates at the 2007 and 2011 elections was 0.51 and 0.20% respectively.

7. Analysis conducted by the author using a variation of Bochsler’s geometric function indicates that the level of exaggeration is likely to be approximately 0.08% at most.

8. This represented 19.81% of votes at the 1987 election, 14.44% at the 1995 election, 18.32% at the 1999 election, 45.33% at the 2002 election, and 13.03% at the 2007 election.

9. The disparity between the vote and seat share of independent candidates for the 1961 and 1965 elections was 0.81 and 3.19% respectively.

10. This was done to investigate the extent to which the method of apportionment introduced after the 1980 coup fostered a predominant party system in favour of the AKP after its election in 2002 (Çarkoğlu and Aksen 2019, 49). The method of apportionment introduced automatically allocated one seat to every province and then applied the Jefferson method, inherently benefitting smaller provinces that would not have otherwise had a population large enough to justify the seat allocation they received (Çarkoğlu and Aksen 2019, 49).

Additional information

Notes on contributors

Jack Jacovou

Jack Jacovou is a PhD candidate at Deakin University, Melbourne. He holds a BA(Hons) degree specialising in Politics from the University of New South Wales. His current research examines how Turkey can be considered to have operated as a hybrid regime between 1960 and 2002. His research interests include Turkish political history, hybrid regimes, and electoral studies. Jack has previously published in the Australian and New Zealand Journal of European Studies and in 2018 won the European Union Award for Excellence for the Best Undergraduate Essay.