![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
Are voters more likely to support candidates who are victims of political persecution? I draw on an original survey with embedded experiments deployed in Turkey ahead of the June 2015 General Election to advance a theory linking political victimhood to an electoral advantage. The results suggest that voters primed with information about a candidate’s political imprisonment, on average, report higher ideological affinity with the candidate. In addition, respondents who identify as co-victims are more likely to say they would vote for a candidate who was imprisoned. These findings are significant and hold regardless of which party the candidate belongs to. Respondents presented with a candidate from the incumbent Justice and Development Party also report higher levels of trust and closeness with the candidate who was imprisoned. This pattern is consistent for voters with low trust in the Justice and Development Party leadership and low levels of religiosity, demonstrating that a history of persecution can broaden candidates’ support.
Acknowledgments
Numerous individuals graciously read various drafts of this article and discussed its central arguments. I am particularly grateful to Bethany Albertson, Caitlin Andrews, Lindsay Benstead, Daniel Brinks, Steven Brooke, Matt Buehler, Daniel Corstange, Jake Dizard, Kristin Fabbe, Sharan Grewal, Gülnur Kocapınar, Ellen Lust, Ken Miller, Liz Nugent, Peter Russell, Daniel Weitzel, Kurt Weyland, and Kerem Yıldırım. Earlier versions of this article were presented at the 2019 annual conference of the Middle East Studies Association and the 2016 annual conference of the American Political Science Association, as well as at seminars and workshops at the University of Texas at Austin and Harvard University. Three anonymous reviewers for Mediterranean Politics also provided detailed comments that considerably improved the article. All errors and failings are my own.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1. The 2015 Turkey General Election Survey was conducted by the author and a Turkish colleague during the 38 days leading up to the 7 June 2015 Turkish General Elections. The data for this article, code for running the models, and the survey instrument will be made available upon request.
2. I use the term ‘Islamist’ to describe politicians who believe that Islam as a body of faith has something important to say about how politics and society should be ordered.
3. The null finding for the CHP candidate is discussed in depth in the results section of this article.
4. Within İstanbul, I conducted interviews in three neighbourhoods from Istanbul’s second electoral district – Pnar, Maden and Mecidiyeköy. Economic development levels in Pnar and Maden were very close to İstanbul’s average. The residents I interviewed in these two neighbourhoods commonly lived in illegally built squatter houses (gecekondu) and worked unskilled, manual jobs. Mecidiyeköy, by contrast, is a prominent business quarter where residents lived in apartment buildings and generally held white-collar occupations. Within each of these neighbourhoods, I randomly selected ten streets from a comprehensive list of all the streets in the neighbourhoods. Household visits were conducted with randomly selected individuals on each of these ten streets. The remaining 20 interviews took place during short research visits to Ankara, Kayseri, Mardin and Diyarbakr. In Ankara and Kayseri, I solicited interviews at local malls. In Diyarbakr and Mardin, I met locals and used snowball sampling to target a predominantly Kurdish sample. I recorded interviews with an audio recorder, when granted consent. An average interview took roughly 1.5 hours, with a minimum of 30 minutes and a maximum of 2.5 hours. Interviews included questions on social characteristics, economic evaluations, religion and religious practice, partisanship, political involvement, evaluation of political parties, assessments of the political situation at the local and national level, and accounts of past and current voting behaviour. Interviews were conducted in Turkish with a native speaker present in most cases to translate anything immediately unclear.
5. The perceived Islamist advantage is rooted in the recent strong performance of political Islam in competitive elections across the region. In the past two decades, Islamist parties have come in first in elections in Turkey (2002-present), Tunisia (2011, 2019), Morocco (2011, 2016), Egypt (2011, 2012), Kuwait (2012, 2016), and Sudan (2015).
6. Author interview, 32-year-old male İstanbul resident, 5 February 2015.
7. Author interview, 50-year-old male İstanbul resident, 5 February 2015.
8. ‘Turkey Bans Islamist Party.’ The Guardian. 17 January 1998.
9. The PKK is considered a terrorist organization in Turkey and several other countries.
10. Author interview, 48-year-old female non-Kurdish İstanbul resident, March 23 2015.
11. Author interview, 55-year-old male non-Kurdish İstanbul resident, March 23 2015.
12. Author interview, 23-year-old female non-Kurdish resident of Ankara, 14 April 2015.
13. Author interview, 22-year-old ethnic Kurdish male resident of Ankara, 12 April 2015. The interviewee’s parents are Kurdish, but he does not traditionally support the Kurdish political parties.
14. Schwedler and others argued that the inclusion of Islamists into politics can lead to their moderation (Cavatorta & Merone, Citation2013; Schwedler, Citation2007; Tezcür, Citation2010). This inclusion-moderation hypothesis concludes that Islamists moderate both to appeal to a larger constituency and as a result of interacting and working with other parties and individuals of vastly different viewpoints on university campuses, civil society, and electoral coalitions.
15. ‘Turkey’s jailed presidential candidate Demirtaş appeals to public via phone call to his wife,’ Stockholm Freedom Centre, 6 June 2018.
16. ‘Turkey’s Opposition Leader’s Arrest Feared.’ Vox Europe. 14 August 2017.
18. http://www.allin1social.com/facebook-statistics/countries?period=three_months.com, http://www.statista.com/statistics/278435/percentage-of-selected-countries-internet-users-on-facebook/.com.
19. Roughly 22% of the TGES sample intended to vote for the pro-Kurdish HDP in June 2015. In contrast, the party’s actual vote share in June 2015 was 13.12%.
20. tests do not indicate imbalance across gender (
(1) = .70, p = .40) or income
(9) = 4.66, p = .86); also demonstrates no more imbalance across pro-Sharia, pro-CHP, pro-AKP, college educated, unemployed, Kurdish, female, and poor segments of the population than we would expect by chance. The TGES sample includes respondents from 200 of the 957 districts in Turkey.
21. Mehmet Yılmaz is a generic name in Turkish, similar to John Smith in English.
22. A limitation of this study is that I was testing support for the hypothesized candidate during a general election, when the electorate tends to cast votes along party lines. Whereas Turkish voters pay close attention to specific candidate characteristics during local elections where they elect mayors and other local representative, they are less inclined to do so when they elect representatives to parliament. I nevertheless focus the study on a candidate for parliament because the survey was conducted ahead of a general election. Were I to frame the politician as a candidate for mayor, it may have been confusing.
23. Questions to gauge political attitudes towards and propensity to vote for the candidate were asked immediately after the respondents read the candidate biography.
24. Reported results reflect a one-tailed test in the expected direction, given my theory and prevailing observational evidence that candidate acts of political suffering increase voter trust, ideological closeness, and propensity to vote for the candidate.
25. Lower numbers of ideological affinity show higher affinity between the candidate and respondent.
26. Respondents are coded as victims if they either ‘strongly disagreed,’ ‘disagreed,’ or ‘somewhat disagreed’ with the idea that ‘secular people in Turkey can live their lives freely’ or supported the ethnic Kurdish political party.
27. Political knowledge is measured from responses to the following question: ‘In which year was the ruling Justice and Development Party (AKP) founded?’ The 24.17% of respondents who answered the question incorrectly are coded as having low political knowledge.
28. ‘Turkey’s Hapless Opposition,’ Foreign Policy, 4 December 2015.