An analysis of pre-election violence through decision theory, experimental design and spatial econometrics

This dissertation explores the multifaceted dynamics of pre-election violence across three interconnected chapters. In the first chapter, I focus on the conflicting perspectives in the literature surrounding the geography of pre-election violence. The proposed decision-theoretic model suggests that the location of violence is largely determined by the perpetrating party’s ability to distinguish its own supporters from those of its rivals, a capability I define as “discrimination capacity.” The findings posit that violence can occur both in competitive districts and party strongholds in settings where perpetrating parties have high discrimination capacity. In contrast, the absence or low levels of this capacity create a stricter geographical divide, with violence concentrated in opposition-dominated districts. The second chapter delves into the impact of pre-election violence on voter behavior. This chapter utilizes a causal identification strategy by leveraging an unexpected inter-party violent event that divided a public opinion survey into two comparable (pre- and post-violence) samples in Zimbabwe. Through a Difference-in-Differences analysis, the chapter reveals differential effects of violence on opposition and government supporters’ mobilization. The third chapter uncovers divergent effects across geographic contexts, emphasizing the role of spatial dynamics in shaping the relationship between heightened levels of fear and political participation within different religious groups in Nigeria. Findings from the geographically weighted regression analysis show that extreme fear leads to political mobilization for Christians and Muslims in areas where they form the majority and demobilization when they are a minority within their surrounding communities. In conclusion, the dissertation findings underscore the importance of governing parties’ discrimination capacity, contextual factors, and spatial heterogeneity in understanding the complex interplay between pre-election violence, extreme fear and voter turnout. The insights presented hold implications for scholars, policymakers, and electoral practitioners navigating the challenges posed by pre-election violence in diverse socio-political contexts.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Political Science

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2023

Author dc:creator

• Dilgin, Tolgahan

Contributors dc:contributor

• Tam Cho, Wendy K.
• Gaines, Brian J
• Hummel, Sarah
• Carroll, Robert

• Pre-election Violence
• Decision Theory
• Natural Experiment
• Geographically Weighted Regression
• Spatial Econometrics

Statement dc:rights

• Copyright 2023 Tolgahan Dilgin

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/122032

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F122032