| Summary: | Submitted in fulfillment of the academic requirements towards a degree in Doctor of Philosophy (PhD) in International Relations, Faculty of Humanities, University of the Witwatersrand, December 2017. === Election violence has been identified as one of the new sources of conflict in Africa, demonstrating resilience rather than transience. Lack of theorization about election violence means that it remains poorly understood as a phenomenon which has consequences in terms of policy formulation and the more practical task of ‘prediction and prevention’. Due to the number of multivariate factors which are present and interact to produce electoral violence, previous attempts at theorizing often result in the problem of equifinality. This study examines the historical and contemporaneous conditions and causes of election violence in Kenya, Nigeria and Zimbabwe from Independence to 2008, a coinciding and ascendant violent election year. The study attempts to locate, identify and trace conditions and causes of election violence, and establish a trajectory and sequence of events that produce election violence, over time, across the three case studies. Comparative case study research of this kind allows for a bounded and in-depth study of causal inference of the select phenomenon over time. A central research question that guides this study is, why do some elections display recurrent violence? The qualitative study takes a historical ‘past-in-the-present’ approach to understanding a manifest contemporary problem by using archival sources, election observer reports, and elite interviewing. Thematic content analysis of themes derived such as ‘election/vote’ ‘violence/protest/uprising/riot’ ‘rigging/fraud’ ‘corruption/vote buying/patronage’ ‘ethnicity’ ‘youth’ ‘land’ ‘resources’ ‘youth wing/militia/gang’, guide the gathering and analysis of the data. The main findings of the research are threefold. Firstly, election violence results predominantly due to elite factionalism and elite disputes whereby elections (process, institutions, campaign and supporters) are politicized. This may take the form of political elites declaring themselves the winner ahead of results being announced; political elites inciting/arming/stoking violence via youth wings, militias, or amongst supporter branches; political elites utilizing inflammatory rhetoric during electoral campaigns; and/or political elites disputing the electoral process, institutions or the final result. Elites are thus a significant factor in election violence. Secondly, election violence results whereby the electoral process and associated state and electoral institutions have been compromised or interfered with such as the political appointment of national electoral commission chairs, members and judges; the partisan deployment of the police and security agencies to repress/intimidate/hinder political competition; the ‘tweaking’ of election results or voter turnout; the enactment of restrictive laws; hegemony of state media; and the allotment of state resources that favour partisan/incumbent sources. Thirdly, election violence results whereby patronage and/or corruption intersect and interact with elites and institutions in producing a violent outcome. This may take the form of buying off/rewarding voters, buying off/rewarding youth militias or violent outfits; buying off/rewarding political patrons. Only certain conditions are significant in the causal change in producing violent electoral outcomes, some such as ethnicity or identity, and land demonstrate durability across the case studies, while others such as unemployment or hunger or inflation are periodically germane. The outlook on election violence in Africa is bleak: in 2017, 6 out of 11 African states that have held elections thus far have been marred by some level of violence. The need to ‘predict and prevent’ the proliferation of this new form of conflict, and work against it becoming a norm, is pressing. This necessarily demands theorizing, knowledge building, and establishing where along the causal chain election violence is likely to break out so that targeted and lasting interventions can occur. === MC2019
|
|---|
