Introduction

• Electoral violence: a sub-type of political violence and electoral fraud, defined as forms of coercive force directed towards electoral actors and/or objects during electoral processes.
• Regional dimension of electoral violence has always been the focus of the relevant literature, but a systematic research on Turkey is non-existent.
• Although the “Kurdish question” has multiple dimensions research mostly concentrated on the economic aspects.
• However, uneven development also reveals itself political processes, not excluding the electoral politics: Especially during the last decade, reviews of the voting process have brought forth a slew of complaints, including reports on voter intimidation and violent attacks on candidates and electoral materials.

Literature

• Aggregate-level factors associated with the phenomenon: regime type, ethnic conflict, competitiveness, the presence of observers and informal and formal institutions (e.g. Asunka et al 2017; Bhasin and Gandhi 2013; Burchard 2015; Collier and Vicente 2012; 2014; Daxecker 2012; 2014; Hafner-Burton, Hyde and Jablonski 2014; van Ham and Lindberg 2015; Höglund and Fjelde 2016; Klaus and Mitchell 2015; Salehyan and Linebarger 2015; Wilkinson 2004)
• Little research on individual-level perceptions of electoral violence and very little research on attribution of blame for violent acts: poor and otherwise under-privileged groups are most likely to fear being targeted by electoral violence (Bratton 2008; Burchard 2015; 2018; Dercon and Gutiérrez-Romero 2012; Gonzalez-Ocantos et al. 2018; Mares and Young 2016).

Literature

• Some indication of a relationship between strength of partisanship and fear of violence; Kenyan case that strong partisans are more likely to fear being targeted by violence (Guttierez-Romero and LeBas 2018).
• Finally, there is evidence to suggest that fear of electoral violence is associated with lower levels of trust and democratic satisfaction, at least in the African context (Burchard 2015; Dercon and Gutierrez-Romero 2012); and little indication that under most conditions fear of electoral violence enhances levels of political knowledge, as would be predicted by Affective Intelligence Theory (Söderström 2017).

Argument

• Our argument is that two factors play important roles in conditioning attributions of blame for violent acts at election time: partisanship and context.
• Partisanship: The role of political party elites in ‘mobilizing bias’ (Schattshneider 1975), bias against outgroups is a common factor in political competition and a frequent determinant of party affiliation in multiethnic societies (Horowitz 1985; Tilly 2004).
• Context: The regional dimension of electoral violence focusing on uneven state capacity and context of ethnic strife, where multi-ethnic grouse where there have historically been tensions over land or other rights are more prone to violence (Alesina, Piccolo and Pinotti 2016; Harish and Toha 2017; Mignozetti and Sexton 2018; Murray 2016; Reno 2011).

Research Questions

Trying to analyse, understand and explain the above-described condition, this study sets two research questions:

RQ1: What is the nature of electoral violence within the regions where the Kurdish population is dense? Are there any differences between different regions of Turkey with regards to electoral violence?

RQ2: How does electoral violence relate with other factors? Does it have unique characteristics belonging to Kurdish regions?

EI-T and DIEV-T

• Two original data sets: EI-T and DIEV-T
• DIEV-T: An incident level database, compiled by a comprehensive content analysis code book, covers electoral violence news which have happened in each and every election that have taken place since 1950 in Turkey.
• EI-T: The survey research on electoral integrity in Turkey (EI-T), includes field experiments and cross-sectional data, measures perceptions on electoral violence and relates the phenomena with other factors including party identification, region, media consumption and trust.

DIEV-T on RQ1: Distributions

Distribution of Electoral Problem Types According to Regions

DIEV-T on RQ1: Distributions

DIEV-T on RQ1: Distributions

DIEV-T on RQ1: Correspondance Analysis on Problems

DIEV-T on RQ1: Correspondance Analysis on Violence

EI-T on RQ2: Models on HDP Voter and ESEA