Chapter 3 summarizes existing theories of election-related violence and introduces a theory of violence as a result of elite misperception. Existing theories rely upon the assumption that politicians choice of violence as an electoral tactic is based on an accurate assessment its relative costs and benefits. The theory of elite misperception, on the other hand, builds on two insights: (1) that voter backlash against violence is substantial and can cost politicians votes and (2) politicians frequently misjudge voter preferences – including with respect to violence – which can lead them to overestimate its benefits as an electoral tactic. The theory points to the difficulties politicians face in accurately assessing the impact of electoral tactics, as well as to the prevalence of cognitive biases that can cause them to misinterpret what information they have. It also highlights the importance of founding elections in determining which electoral tactics – violence included – are considered effective and worthy of emulation. With these insights in mind, the theory explains the incidence and persistence of election-related violence as a result of politicians misperceptions about voter preferences and the effects of violence on voting rather than the objective electoral benefits that violence provides.

Chapter 2 summarizes the nature and extent of election-related violence globally and in Kenya specially. It documents the various forms that such violence takes and provides background on the Kenyan case, noting certain key features that make it particularly useful to study. In particular, it establishes that Kenya is a case where (1) political elites play a primary role in instigating violence and (2) elections are competitive enough that voters have a genuine choice at the polls. It also establishes it as a hard case for testing a theory of elite misperception, as the conventional wisdom holds that violence – working through several of the mechanisms posited in the literature – is an effective tool for winning Kenyan elections. It concludes with some discussion of the special role that ethnicity often plays in the outbreak of violence in electoral competition.

Chapter 6 addressed the puzzle of why politicians employ violence as an electoral tactic in Kenya when the benefits of doing so are uncertain at best. Data from survey experiments with politicians that parallel those conducted with voters – as well as evidence from qualitative interviews – show that, contrary to what the literature assumes, politicians misperceive the effects of violence and violent ethnic rhetoric on voter preferences over candidates for office, underestimating the size and breadth of voter backlash against the use of these tactics. This misperception explains why election-related violence continues to occur in Kenya despite its questionable efficacy as an electoral tactic. Furthermore, access to information alone does not appear to be enough to correct politicians misperceptions in this domain. Elite misperception can explain why violence occurs in the course of electoral competition even when its efficacy is in doubt.

This chapter summarizes the argument and findings presented in the book, explores a number of their implications, and discusses their relevance to broader debates. It argues that the books findings point to a need for research to more carefully evaluate the costs of violent electoral tactics in addition to its electoral benefits, including more micro-level research – such as that presented in the book – that explores voter responses to violence. In addition, scholars should ask more explicitly whether and how political elites are able to accurately assess the relative costs and benefits of violence and other electoral tactics. Future research should delve deeper into the question of how and why elite misperceive voter preferences, and when and why it is most likely to occur.

Chapter 4 evaluates the overall relationship between violence and election outcomes, finding that violence provides no benefit, on average, to the parties that use it. In some cases, it may even undermine the candidacies of the politicians associated with it. In particular, an analysis of the relationship between the incidence of violence prior to elections in Kenyas first two multiparty elections in the 1990s and constituency-level election outcomes for the ruling KANU party (the primary instigator of the violence) finds no relationship between the incidence of violence and KANU vote share or the likelihood that their parliamentary candidates won election. Second, an analysis of the subsequent electoral performance of candidates allegedly involved in the large outbreak of violence in 2007/08 finds that these candidates lost their reelection bids at a much higher rate than the average incumbent. Several lost in rather unusual ways, including by losing their party primaries, losing to minority ethnic group candidates, or losing the general election after securing the nomination of the locally dominant party.

This chapter introduces the research question and motivation for the book. It introduces and defines the concept of election-related violence, defining it as physical acts of violence carried out for the purpose of affecting the results of an election. It introduces existing explanations for election violence and previews the theory of elite misperception presented in the book, highlighting its basis in observations about voters opinions about violence and politicians difficultly in inferring what voters want. Finally, it summarizes the empirical strategy the book uses to evaluate the argument and competing explanations, and it provides a roadmap for the rest of the book.

This chapter explores whether and how the findings from Kenya travel elsewhere around the world, analyzing cross-national quantitative analyses and case studies of a diverse group of countries with historically varying levels of violence, including Indonesia, Pakistan, Ghana, Nigeria, and Brazil. The analyses show how the causes and consequences of election-related violence identified in Kenya can help explain violence (or its absence) in other competitive electoral contexts as well. The cross-national quantitative analysis demonstrates the key role that the conduct of founding elections – in particular, the tactics used by the founding election winner – appears to play in whether or not violence becomes commonly employed by parties and candidates in the electoral cycles that follow. The case studies reveal little to no evidence that violence is effective in helping parties and candidates win office but do identify instances in which it may have resulted in voter backlash. In addition to identifying key structural conditions that make violence a more viable option in some cases than others, the analysis finds evidence consistent with the idea that the conduct of founding elections plays an important role in shaping politicians beliefs and expectations about the effects of violence on their electoral prospects.

This chapter analyzes the effects of violence – via coercion and persuasion – on voting. Existing research provides some evidence that violence can be used to prevent voters from turning out to vote. However, an analysis of the relationship between violence and turnout in Kenyan elections in the 1990s finds no relationship between violence on the one hand, and the aggregate level of voter registration or turnout on the other, suggesting that the relationship between violence, turn-out and electoral geography may be less straightforward than previously thought. On persuasion, the analysis finds consistent evidence of a large and broad-based voter backlash against violence, including among the coethnic voters that politicians rely on to win elections. Furthermore, while antagonistic ethnic rhetoric does appear to increase the likelihood of violence breaking out, it is not a useful strategy for mobilizing coethnic support. These results suggest that voter backlash against violence and violent rhetoric may undermine the efficacy of violence as an electoral tactic, and they help explain why violence is not associated with better outcomes for candidates in real elections.

Chapter 5 turns to the strategy of violence outsourcing by examining a party that shared many of the same incentives in Karachi as the Muttahida Qaumi Movement for carrying out violence but which did not possess the capacity to do so itself. I focus on how the Pakistan People’s Party (PPP) relied on the People’s Aman Committee (PAC), an ethnic militia in Karachi, both to meet the demands of its constituents in the neighborhood of Lyari and to carry out the “dirty work” of violence, intimidation, and extortion. Through the PAC, the PPP was able to engage in voter fraud and intimidation; engage in criminal activity and ‘turf wars’ with the MQM and ANP over valuable economic property; and maintain support among its co-ethnic base by further polarizing the electorate along ethnic lines. In Karachi’s polarized environment, the PPP had a generally captive support base of Sindhi and Baloch voters in the city who imposed minimal electoral costs on the party for violence given their lack of alternative options. Despite evident principal–agent problems, the PPP relied on the gang because it did not have the local-level organizational capacity to carry out violence – or gather votes – itself.

Chapter 4 examines the phenomenon of direct party violence, examining why and how the Muttahida Qaumi Movement (MQM) engaged in violence in Karachi between 1986 and 2016. I show that the MQM was able to reap the numerous benefits of violence in the ethnically polarized, Hobbesian landscape of Karachi without losing the support of its core, Muhajir constituency. The MQM maintained a captive support base among the plurality Muhajir ethnic group, who perceived few alternative options available for purposes of political representation and who did not therefore punish the party electorally for its involvement in violence. Survey experimental results based on an original conjoint survey are striking: The likelihood of Muhajirs supporting a violent MQM candidate is identical to the probability of their supporting a peaceful MQM candidate. Relying on my own qualitative fieldwork and secondary ethnographic accounts, I show that the MQM used its own militant cadres to target the opposition and engage in turf wars with rival ethnic groups. It was able to do so because it was an organizationally strong political party with committed and socialized party workers willing to engage in risky action.

Chapter 6 analyzes the escalation and production of electoral violence, with a focus on Kenya’s 2007–2008 postelection. The main question asks how land narratives shape the escalation and occurrence of violence. Using a process-tracing approach based on qualitative evidence from three communities in Nakuru County, the chapter demonstrates how land narratives contribute to the production of election violence. I take the reader through four stages in the escalation of electoral violence. These stages include: (1) the mobilization of land narratives, (2) the trigger event, (3) local escalation, and (4) scale-shift. As part of this analysis, the chapter explains how different logics of violence, from preemption, opportunities to alter the status quo, revenge, and desire for material gain play out at these different stages. Broadly, the chapter shows how local land narratives can provide a key discursive tool through which both elites and ordinary civilians establish motives for organizing and engaging in violence.

Chapter 1, the introduction, begins by motivating the puzzle of election violence, asking why election violence escalates in one local context, but not another seemingly similar context. In contrast to existing accounts that emphasize institutional features of the state or the strategic incentives of political elites, the chapter introduces the book’s main argument: that the occurrence of violence is a joint production between political elites and ordinary citizens. In outlining this argument, the chapter introduces the book’s theory of land narratives and their role in the production and process of electoral violence, arguing that land narratives can serve as a key device around which elites and citizens coordinate the use of violence. The chapter then summarizes the book’s methodology, reviews existing explanations of electoral violence, and provides an outline of the book.

Chapter 7 examines why contentious land narratives are not sufficient predictors of electoral violence. In contrast to the previous chapter, which demonstrates how elites use narratives to organize violence, this chapter draws on evidence from counties in the Coast region where there are salient contentious land narratives yet electoral violence is rare. The chapter argues that land narratives work differently along the Coast because residents do not link their land rights with electoral outcomes. Hence, residents have few motives to participate in electoral violence and politicians have far less power to use land narratives to organize violence. To account for this regional difference, the chapter brings the reader back to the theory of “landlord” and “land patron,” which it discusses in terms of patronage strength. It also explains the importance of group size: the proportion of ethnic insiders relative to outsiders at the local level.

Chapter 2 presents the book’s main theory, which links contests over land with the local-level dynamics and escalation of electoral violence. It argues that the escalation of electoral violence is part of an historically-rooted process that includes inequality in land rights between two identity-based groups, the formation of contentious land narratives between these groups, and the mobilization of these land narrative to organize and produce electoral violence. The chapter explains each of these “stages” in the process of violence. It begins by theorizing how inequality in land rights between groups can shape a distinct set of contentious narratives around land. It then explains how these land narratives can shape political action, enabling or restraining the production of violence. The chapter also explains the research design and methodology that provides the evidentiary basis for this theory of electoral violence.

This chapter shifts from the more inductive approach that guides preceding chapters to a deductive one, using survey data to test existing theories about the causes and consequences of electoral violence. In doing so, the chapter shifts the unit of analysis from the region and group-level to the individual. The chapter has two main parts. The first examines the predictors of electoral violence, focusing specifically on the role of divisive land appeals in increasing an individual’s likelihood of experiencing violence. The second part focuses on the effects of violence, asking how the experience of election violence shapes openness toward ethnic outgroups, trust in political leadership, and engagement across ethnic lines. Broadly, the chapter shows that the experience of election violence has an enduring effect on how an individual perceives and engages with her political and social world. The chapter also emphasizes that studying the effects of electoral violence helps unpack the potential endogeneity of violence, enabling scholars to better specify the mechanisms through which election violence increases or diminishes the prospects for democratic consolidation and durable peace.

Chapter 9, the conclusion, provides a brief summary of the book’s main argument and discusses its generalizability. To do so, it draws on examples that illustrate how the book’s theory might generalize to other cases such as Burundi, Côte d’Ivoire, and the Democratic Republic of the Congo, as well as cases such as Colombia, Peru, and India. The chapter then discusses how the main argument can inform broader theorizing about the relationship between democratization, political violence, and postconflict politics. Finally, it reflects on potential lessons learned from a close study of land rights, group narratives, and violence.